Pressure sensor real-time fault detection method based on dynamic environment adaptation

Abstract

The invention provides a pressure sensor real-time fault detection method based on dynamic environment adaptation, and belongs to the technical field of sensor fault detection. The method comprises the following steps: establishing a time domain association data set of noise-sensor signals by synchronously collecting environmental noise signals and pressure sensor output signals in real time; performing spectral analysis on the environmental noise signal, dynamically dividing frequency bands, and calculating energy distribution of each frequency band; dynamically generating an operation delay threshold according to the energy distribution and the weight factor of each frequency band; calculating the reliability coefficient of the operation state of the sensor by combining the time domain characteristics of the output signal of the pressure sensor and the current noise energy distribution; when it is detected that the output signal of the pressure sensor is abnormal, the abnormal duration is corrected based on the operation state credibility coefficient, the corrected duration is compared with the dynamic operation delay threshold value, and whether the fault is a real fault or not is judged. According to the invention, environmental noise interference can be effectively suppressed, and the accuracy and reliability of pressure sensor fault detection are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of sensor fault detection, and more particularly, to a real-time fault detection method for pressure sensors based on dynamic environment adaptation. Background Art

[0002] As a core sensing element in industrial automation systems, the measurement accuracy and reliability of pressure sensors directly affect the stability and safety of the production process. In complex industrial environments, pressure sensors are long-term exposed to dynamic environmental factors such as high-frequency noise interference, temperature and humidity fluctuations. These interferences not only mask the true output signal of the sensor, but also cause the accumulation of measurement errors and even false alarms. Traditional fault detection methods often struggle to distinguish environmental noise interference from real sensor faults, posing severe challenges to the accuracy and timeliness of fault diagnosis.

[0003] The pressure sensor fault detection system disclosed in Patent CN116878728B represents a typical current technical solution, which judges environmental interference by setting a fixed noise threshold. Although this system first incorporates environmental noise monitoring into the sensor fault detection system, its core adopts a static threshold comparison mechanism: when the noise value collected by the environmental noise detection module exceeds a preset fixed threshold, it is determined that the sensor may be interfered. This rigid judgment method has a certain effect in simple working conditions, but it cannot cope with the complex noise dynamic change characteristics in industrial sites.

[0004] Since this technology uses a static noise threshold as the sole judgment criterion, the system has fundamental defects: on the one hand, the fixed threshold cannot adapt to the dynamic changes in noise intensity under different working conditions, and a large number of false judgments are bound to occur in scenarios with severe noise fluctuations; on the other hand, simply relying on noise amplitude comparison while ignoring noise spectrum feature analysis makes the system unable to distinguish the essential differences between high-frequency abnormal noise and normal working condition noise. This defect directly causes two serious consequences: 1) when the environmental noise fluctuates naturally, the system frequently misreports faults; 2) when a real sensor fault is accompanied by noise in a specific frequency band, the system may miss the report. Summary of the Invention

[0005] In order to solve the problems of lagging fault detection caused by environmental noise interference and static analysis in the prior art, and the disconnection between environmental noise and sensor operating state analysis in the prior art, and to effectively distinguish the interference between high-frequency abnormal noise and normal working condition noise, and improve the timeliness and accuracy of fault detection, the present invention provides a real-time fault detection method for pressure sensors based on dynamic environment adaptation.

[0006] To this end, the present invention adopts the following technical solutions: On the one hand, the present invention discloses a real-time fault detection method for pressure sensors based on dynamic environment adaptation, including the following steps: Collect the environmental noise signal and the output signal of the pressure sensor in real-time synchronization, and establish a time-domain correlation dataset of the noise-sensor signal; Conduct spectral analysis on the environmental noise signal, dynamically divide the noise spectrum into low-frequency band, medium-frequency band and high-frequency band, and calculate the energy distribution of each band; Dynamically generate an operation delay threshold according to the energy distribution of each band and the weight factor, where the weight of the high-frequency band energy in threshold adjustment is higher than that of the low-frequency band and the medium-frequency band; Combine the time-domain characteristics of the output signal of the pressure sensor and the current noise energy distribution to calculate the reliability coefficient of the sensor operation state; When it is detected that the output signal of the pressure sensor is abnormal, correct the abnormal duration based on the reliability coefficient of the operation state, and compare the corrected duration with the dynamic operation delay threshold to determine whether it is a real fault.

[0007] Further, collecting the environmental noise signal and the output signal of the pressure sensor in real-time synchronization includes: Synchronously start the data acquisition of the noise sensor and the pressure sensor by using a hardware trigger signal; The length of the sampling time window is dynamically adjusted according to the working frequency of the pressure sensor, and the adjustment range is 10ms - 200ms.

[0008] Further, the dynamic division of the noise spectrum includes: Obtain the spectral distribution of the noise signal through fast Fourier transform; Use an adaptive clustering algorithm to divide the spectrum into low-frequency band, medium-frequency band and high-frequency band, and the division boundary is dynamically adjusted according to the environmental noise characteristics.

[0009] Further, the low-frequency band is 0 - 500Hz, the medium-frequency band is 500Hz - 2000Hz, and the high-frequency band is above 2000Hz.

[0010] Further, dynamically generating an operation delay threshold according to the energy distribution of each band and the weight factor includes: Adjust the weight coefficients of each band according to the current working stage of the sensor; Generate a dynamic delay threshold based on the following formula : ; where: is a preset reference threshold; is the high-frequency energy activation threshold; is the slope coefficient; is the total noise energy; , , are the proportion of low-frequency, medium-frequency and high-frequency noise energy; , , respectively represent the dynamic weight coefficients of the low frequency band, the middle frequency band and the high frequency band.

[0011] Furthermore, the dynamic weight coefficients , , satisfy: ; Adjust the weight coefficients of each frequency band according to the current working stage of the sensor, including: Startup stage: the high frequency weight increases by 20% - 30%; Stable stage: the middle frequency weight is increased to 1.2 times the reference value; Shutdown stage: the low frequency weight decays exponentially.

[0012] Furthermore, the calculation of the running state credibility coefficient includes: ; where the time domain credibility factor The calculation formula is: ; the frequency domain credibility factor The calculation formula is: ; In the formula, , , are the low frequency, middle frequency and high frequency noise energy ratios; is the total noise energy; is the abnormal duration, reflecting the time window for fault detection; represents the current moment, that is, the time point for calculating the credibility; is the integration variable, representing a certain time point within the sliding time window, ranging from to ; is the detection signal of the pressure sensor at moment; , are the signal statistics within the sliding time window; , are the steepness coefficients of the sigmoid function; , are the energy ratio thresholds; is the time decay factor, is the credible weight coefficient of the high frequency band.

[0013] Furthermore, correct the abnormal duration based on the running state credibility coefficient, including: Establish a credibility-time correction mapping function: ; where: is the corrected abnormal duration; is the initially set abnormal duration; is the reference credibility threshold; is the correction intensity coefficient; obtains the corrected abnormal duration based on the operating state credibility coefficient; uses the exponential smoothing algorithm to optimize the corrected abnormal duration: ; where: is the optimized abnormal duration, is the smoothing factor; is the abnormal duration after the previous correction.

[0014] On the other hand, the present invention also provides a pressure sensor system, including: a pressure sensing unit; an environmental noise monitoring unit; a fault detection module that performs fault detection using the above-mentioned fault detection method; and an alarm output interface.

[0015] Compared with the prior art, the present invention has the following beneficial effects: Through the technical means of dynamic spectrum analysis and adaptive frequency band division, the present invention monitors the spectrum characteristics of environmental noise in real time and dynamically adjusts the detection frequency band boundary, achieving the technical effect of accurately identifying the noise interference characteristics in different frequency bands; through the technical means of joint acquisition of noise-sensor signals and time-domain correlation analysis, the present invention establishes a dynamic correspondence relationship between noise and sensor signals, achieving the technical effects of eliminating signal acquisition delay errors and improving signal correlation; through the technical means of adaptive threshold generation based on multi-band energy distribution, the present invention dynamically adjusts the fault determination threshold according to the noise spectrum characteristics, achieving the technical effect of maintaining stable detection performance in a complex noise environment; through the technical means of credibility coefficient calculation and abnormal duration correction, the present invention comprehensively evaluates the time-domain signal characteristics and frequency-domain noise distribution, achieving the technical effect of accurately distinguishing real faults from environmental interference; through the technical means of multi-dimensional signal fusion and secondary verification, the present invention combines historical data comparison and comparison analysis of sensors in the same batch, achieving the technical effect of improving the reliability of fault determination. The systematic application of these technical means enables the pressure sensor fault detection system to adapt to the changes in noise interference in a complex industrial environment, significantly improving the accuracy and reliability of fault detection, while effectively reducing the false alarm rate and missed alarm rate, providing a more stable and reliable sensor monitoring guarantee for industrial automation systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1Flowchart of a real-time fault detection method for a pressure sensor based on dynamic environment adaptation and multi-dimensional fusion according to the present invention; Figure 2 Block diagram of a pressure sensor system according to the present invention. Detailed implementation manners

[0018] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below 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 of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0019] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0020] As Figure 1 shown, a real-time fault detection method for a pressure sensor based on dynamic environment adaptation includes the following steps: S1: Synchronously collect the environmental noise signal and the output signal of the pressure sensor in real time, and establish a time-domain correlation data set of the noise-sensor signal; Specifically, the data collection of the noise sensor and the pressure sensor is synchronously started by using a hardware trigger signal. The system sets a unified hardware trigger signal source, and this signal source is simultaneously connected to the acquisition channels of the noise sensor and the pressure sensor through a splitter to ensure that the acquisition start times of the two signals are completely synchronized, and eliminate the correlation analysis error caused by the time deviation.

[0021] The length of the sampling time window is dynamically adjusted according to the operating frequency of the pressure sensor, and the adjustment range is 10 ms - 200 ms. For pressure sensors operating at high frequencies, the system automatically sets a shorter sampling window, such as 10 ms - 50 ms, to capture the characteristics of rapidly changing signals; for pressure sensors operating at medium frequencies, the sampling window is set to 50 ms - 100 ms; for pressure sensors operating at low frequencies, the sampling window is set to 100 ms - 200 ms to ensure that a complete signal cycle is collected.

[0022] During the acquisition process, the system pairs the noise signal and the output signal of the pressure sensor according to the time stamp to form a time-domain correlation data set of noise-sensor signals. This data set contains information such as the amplitude and frequency characteristics of the noise signal, as well as the output value and rate of change of the pressure sensor at the corresponding moment, providing a data basis for subsequent analysis.

[0023] S2: Perform spectral analysis on the environmental noise signal, dynamically divide the noise spectrum into low-frequency band, medium-frequency band and high-frequency band, and calculate the energy distribution of each band; Specifically, obtain the spectral distribution of the noise signal through fast Fourier transform. The system applies the fast Fourier transform algorithm to the collected environmental noise signal to convert the time-domain signal into a frequency-domain representation, obtaining the energy distribution of the noise signal at different frequencies. To improve the accuracy of spectral analysis, the system uses a Hanning window function to preprocess the original signal to reduce the spectral leakage phenomenon.

[0024] Use an adaptive clustering algorithm to divide the spectrum into low-frequency band, medium-frequency band and high-frequency band, and the division boundaries are dynamically adjusted according to the characteristics of the environmental noise. The system first calculates the energy distribution curve of the noise spectrum, and then uses the K-means clustering algorithm to automatically identify the main concentrated areas of the energy distribution, and determines the band division boundaries according to the characteristics of the energy density change. Under standard working conditions, the low-frequency band is 0 - 500 Hz, the medium-frequency band is 500 Hz - 2000 Hz, and the high-frequency band is above 2000 Hz. However, under special environmental conditions, the system will dynamically adjust these boundary values according to the actual noise characteristics.

[0025] The system calculates the energy distribution of each band, including the total energy value, energy ratio, peak frequency and its amplitude of each band. For the low-frequency band, the system focuses on the energy distribution in the range of 0 - 100 Hz, which is usually related to mechanical vibration; for the medium-frequency band, the system analyzes the energy change in the range of 500 Hz - 1000 Hz, which often reflects the mechanical operating state in the environment; for the high-frequency band, the system monitors the energy fluctuation in the range of 2000 Hz - 5000 Hz, which may indicate electrical interference or other high-frequency noise sources.

[0026] S3: Dynamically generate an operation delay threshold according to the energy distribution of each frequency band and the weight factor, where the weight of the high-frequency band energy in threshold adjustment is higher than that of the low-frequency band and the medium-frequency band; Specifically, adjust the weight coefficients of each frequency band according to the current working stage of the sensor. The system identifies the working stages of the pressure sensor, including the startup stage, the stable stage, and the shutdown stage, and sets different frequency band weight coefficients for different stages.

[0027] The dynamic weight coefficients of the low-frequency band, the medium-frequency band, and the high-frequency band , , satisfy: ; Under standard working conditions, the reference weight is set to = 0.2, = 0.3, = 0.5, reflecting the importance of high-frequency noise interference to the sensor.

[0028] In the startup stage, the high-frequency weight increases by 20% - 30%, that is, it is increased from the reference value of 0.5 to 0.6 - 0.65. At the same time, the low-frequency and medium-frequency weights are correspondingly reduced to adapt to the possible high-frequency electrical interference during startup. In the stable stage, the medium-frequency weight is increased to 1.2 times the reference value, that is, it is increased from 0.3 to 0.36. At the same time, the high-frequency weight is appropriately reduced to pay attention to the influence of mechanical vibration during stable operation. In the shutdown stage, the low-frequency weight decays exponentially, with an initial value of 0.2 and gradually decreasing to 0.1 over time to reduce the influence of low-frequency vibration on judgment during shutdown.

[0029] The dynamic operation delay threshold is the fault determination time threshold adaptively adjusted by the pressure sensor system according to the real-time working conditions. When the abnormal duration of the pressure sensor signal exceeds this threshold, it is determined as a real fault.

[0030] Generate the dynamic delay threshold based on the following formula : ; where: is the preset reference threshold, with the unit of millisecond, set according to the sensor model and application scenario, usually 50ms - 200ms; is the high-frequency energy activation threshold, set to 1.5 times the high-frequency energy of the environmental reference noise; is the slope coefficient, with a value range of 0.01 - 0.05, used to control the influence degree of noise energy on the threshold; is the total noise energy, obtained by calculating the root mean square value of the collected signal; , , Let \(P_{LF}\), \(P_{MF}\), and \(P_{HF}\) be the proportions of low-frequency, mid-frequency, and high-frequency noise energy, respectively, representing the proportion of energy in each frequency band to the total energy.

[0031] When the high-frequency noise energy exceeds the activation threshold , the system will significantly increase the operating delay threshold to avoid misjudgment caused by high-frequency interference; when the low-frequency noise energy has a relatively high proportion, the threshold adjustment is relatively small to maintain sensitivity to low-frequency anomalies.

[0032] S4: Combine the time-domain characteristics of the output signal of the pressure sensor with the current noise energy distribution to calculate the confidence coefficient of the sensor operating state; Among them, the confidence coefficient of the operating state is a dynamic weight value (usually ranging from 0 to 1), used to quantify the credibility of the current operating state of the pressure sensor; it synthesizes factors such as historical data consistency, environmental conditions, and device health status, and reflects the reliability of sensor signal anomalies.

[0033] Specifically, the calculation of the confidence coefficient of the operating state includes: ; where the time-domain confidence factor is calculated by the formula: ; the frequency-domain confidence factor is calculated by the formula: ; In the formula, , , are the proportions of low-frequency, mid-frequency, and high-frequency noise energy; is the total noise energy; is the abnormal duration, in milliseconds, reflecting the time window for fault detection; represents the current time, that is, the time point for calculating the confidence; is the integration variable, representing a certain time point within the sliding time window, ranging from to ; is the detection signal of the pressure sensor at time; , are the signal statistics within the sliding time window, representing the standard deviation and coefficient of variation of the signal respectively; , are the steepness coefficients of the sigmoid function, set to 0.05 and 0.1 respectively; , are the energy proportion thresholds, set to 0.4 and 100 ms respectively; is the time decay factor, set to 0.01; is the high-frequency band credible weight coefficient, set to 0.6.

[0034] When the proportion of high-frequency noise energy exceeds the preset threshold, the sensitivity to short-time signal anomalies is reduced. The specific implementation method is: when > , the frequency-domain credibility factor value decreases, resulting in a decrease in the overall credibility coefficient , and the system adopts a more conservative judgment strategy for signal anomalies occurring within a short time, requiring the anomaly to last longer before triggering a fault alarm.

[0035] When low-frequency noise energy dominates, the detection weight for continuous abnormal signals is increased. The specific implementation method is: when > 0.5, the system automatically adjusts the time decay factor in the calculation of time-domain credibility, reducing it from the reference value of 0.01 to 0.005, so that continuous abnormal signals obtain a higher weight in credibility evaluation, effectively identifying the slow fault development process in a low-frequency noise environment.

[0036] S5: When an abnormal signal is detected in the output of the pressure sensor, the abnormal duration is corrected based on the operating state credibility coefficient, and the corrected duration is compared with the dynamic operating delay threshold to determine whether it is a real fault.

[0037] Specifically, correcting the abnormal duration based on the operating state credibility coefficient includes: Establish a credibility-time correction mapping function: ; where: is the corrected abnormal duration; is the initially set abnormal duration; is the reference credibility threshold, set to 0.7; is the correction intensity coefficient, set to 2.0, used to control the influence degree of credibility on time correction.

[0038] Obtain the corrected abnormal duration based on the operating state credibility coefficient. When the credibility coefficient is lower than the reference threshold , the system extends the abnormal duration, requiring the abnormal signal to last longer before being determined as a fault; when the credibility coefficient is higher than the reference threshold , the system shortens the abnormal duration to accelerate the fault determination process.

[0039] Optimize the corrected abnormal duration using the exponential smoothing algorithm: ; wherein: is the optimized anomaly duration, is the smoothing factor, set to 0.7; is the anomaly duration after the previous correction; this smoothing process can reduce the mutation of the correction time and improve the stability of fault determination.

[0040] The system compares the optimized anomaly duration with the dynamic operation delay threshold : When > it is determined as a fault signal. At this time, the fault alarm process can be directly triggered, or secondary verification can be performed. After confirming it as a real fault signal, the fault alarm process is triggered; when ≤ it is determined as a false anomaly caused by temporary anomaly or environmental interference, and continuous monitoring is carried out without triggering the fault alarm.

[0041] For the signal determined as a fault, secondary verification is carried out through at least one of the following methods: Check the recurrence frequency of similar anomaly patterns in the historical data of the same sensor. The system extracts the operation data of this sensor in the past 30 days from the database, analyzes the occurrence frequency and duration of similar anomaly patterns. If the similar anomaly has occurred more than 3 times in the past and the duration of each time exceeds 80% of the threshold, it is confirmed as a real fault.

[0042] Compare the output differences of other sensors in the same batch at the same time period. The system automatically selects 3 - 5 sensors of the same model in the same working environment as the reference group, compares the output signal characteristics of them at the same time period. If the anomaly characteristics of the target sensor do not appear or appear significantly lower in the reference group, it is confirmed as a real fault.

[0043] Inject a test signal to verify the response characteristics of the sensor. The system inputs a standard test signal to the sensor suspected of failure, analyzes the characteristics of the response curve of the sensor, including parameters such as response time, amplitude accuracy, and signal stability. If the response characteristics deviate from the normal range by more than 20%, it is confirmed as a real fault.

[0044] The secondary verification is used to verify the fault signal determined by the above - mentioned pressure sensor fault detection method to avoid false alarms. However, since the secondary verification takes time, in practical applications, the secondary verification is mainly applied to the test stage and the initial stage of implementation.

[0045] In another embodiment, as Figure 2 shown, a pressure sensor system includes: The pressure sensing unit 100 is used to measure the pressure change in the working environment and convert it into an electrical signal for output. This unit includes a piezoresistive element, a signal conditioning circuit, and a data conversion module, and can realize the functions of pressure signal acquisition, conversion, and output. The piezoresistive element uses a silicon piezoresistive sensor with a measurement range of 0 - 10 MPa, an accuracy of ±0.1% FS, and a response time of less than 5 ms. The signal conditioning circuit is composed of a low-noise operational amplifier with a signal-to-noise ratio greater than 60 dB. The data conversion module uses a 16-bit ADC with a sampling rate of up to 10 kHz.

[0046] The environmental noise monitoring unit 200 is used to collect the noise signal in the working environment in real time. This unit includes a broadband microphone array, a preamplifier, and a spectrum analysis processor, and can capture the environmental noise in the range of 0 - 10 kHz, with a sensitivity of -40 dB and a dynamic range of 100 dB. The microphone array consists of 4 omnidirectional microphones arranged in a square with a side length of 10 cm, and can realize noise source localization and noise feature extraction.

[0047] The fault detection unit 300 performs fault detection using the fault detection method described in Embodiment 1. The fault detection unit 300 is respectively connected to the pressure sensing unit 100 and the environmental noise monitoring unit 200, and includes a signal processing subunit, a feature extraction subunit, and a decision-making subunit, and can realize the functions of correlation analysis of noise signals and pressure signals, anomaly detection, and fault determination. The signal processing subunit uses a DSP chip to achieve real-time signal processing with an operation speed of 1 GFLOPS. The feature extraction subunit realizes spectrum analysis and time-domain feature extraction, and supports 512-point FFT operations. The decision-making subunit realizes fault determination based on a credibility evaluation algorithm with a determination delay of less than 200 ms. For the specific fault detection method, refer to Embodiment 1 and will not be elaborated here.

[0048] The alarm output interface 400 is used to provide fault alarm information to the upper-level system or users. This interface supports multiple output methods, including relay switch output, 4 - 20 mA analog output, RS485 digital communication output, and wireless communication output. The relay output supports two modes of normally open / normally closed, with a maximum switching current of 2 A. The analog output accuracy is 0.1% FS. The RS485 communication supports the Modbus protocol with a communication rate of up to 115200 bps. The wireless communication supports WiFi and Bluetooth 4.0 protocols, with communication distances of 100 m and 30 m respectively.

[0049] In several embodiments provided by the present application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the division of the units can be a logical function division. In actual implementation, there can be other division methods. 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 coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of units or modules can be in electrical or other forms.

[0050] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0051] In addition, each functional unit in various embodiments of the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0052] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: USB flash drives, read-only memories (ROMs), random access memories (RAMs), mobile hard disks, magnetic disks, or optical discs, and other media that can store program codes.

[0053] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of various embodiments of the present invention.

Claims

1. A real-time fault detection method for a pressure sensor based on dynamic environment adaptation, characterized in that, It includes the following steps: Collect the environmental noise signal and the output signal of the pressure sensor in real-time synchronization, and establish a time-domain correlation data set of the noise-sensor signal; Perform spectral analysis on the environmental noise signal, dynamically divide the noise spectrum into a low-frequency band, a medium-frequency band, and a high-frequency band, and calculate the energy distribution of each band; Dynamically generate an operation delay threshold according to the energy distribution of each band and the weight factor, where the weight of the high-frequency band energy in threshold adjustment is higher than that of the low-frequency band and the medium-frequency band; Calculate the reliability coefficient of the sensor operation state by combining the time-domain characteristics of the output signal of the pressure sensor and the current noise energy distribution; When it is detected that the output signal of the pressure sensor is abnormal, correct the abnormal duration based on the operation state reliability coefficient, and compare the corrected duration with the dynamic operation delay threshold to determine whether it is a real fault.

2. The method according to claim 1, characterized in that, Collect the environmental noise signal and the output signal of the pressure sensor in real-time synchronization, including: Use a hardware trigger signal to synchronously start the data collection of the noise sensor and the pressure sensor; The length of the sampling time window is dynamically adjusted according to the working frequency of the pressure sensor, and the adjustment range is 10ms - 200ms.

3. The method according to claim 1, wherein The dynamic division of the noise spectrum includes: Obtain the spectral distribution of the noise signal through fast Fourier transform; Use an adaptive clustering algorithm to divide the spectrum into a low-frequency band, a medium-frequency band, and a high-frequency band, and the division boundaries are dynamically adjusted according to the characteristics of the environmental noise.

4. The method according to claim 3, characterized in that, The low-frequency band is 0 - 500Hz, the medium-frequency band is 500Hz - 2000Hz, and the high-frequency band is above 2000Hz.

5. The method according to claim 1, characterized in that Dynamically generate an operation delay threshold according to the energy distribution of each band and the weight factor, including: Adjust the weight coefficients of each band according to the current working stage of the sensor; Generate a dynamic delay threshold based on the following formula :[[]]END]] ； Wherein: is a preset reference threshold; is a high-frequency energy activation threshold; is a slope coefficient; is the total noise energy; 、 、 are the proportions of low-frequency, medium-frequency, and high-frequency noise energy; 、 、 respectively represent the dynamic weight coefficients of the low-frequency band, the medium-frequency band, and the high-frequency band.

6. The method according to claim 5, characterized in that, Dynamic weight coefficients for the low-frequency band, mid-frequency band, and high-frequency band , , Satisfy: ; Adjust the weight coefficients of each band according to the current working stage of the sensor, including: Startup phase: High-frequency weights Increase by 20% - 30%; Stable stage: Intermediate frequency weight Increase to 1.2 times the reference value; Shutdown phase: Low-frequency weight Decay exponentially.

7. The method according to claim 1, wherein The calculation of the running state credibility coefficient includes: ; Among them, the time-domain credibility factor The calculation formula is as follows: ； Frequency domain credibility factor The calculation formula is as follows: ； In the formula, , , are the proportion of low-frequency, medium-frequency, and high-frequency noise energy; is the total noise energy; is the abnormal duration, reflecting the time window for fault detection; represents the current moment, i.e., the time point for calculating the credibility; is an integration variable, representing a certain time point within the sliding time window, ranging from to ; is the detection signal of the pressure sensor at moment; , are the signal statistics within the sliding time window; , are the steepness coefficients of the sigmoid function; , are the energy ratio thresholds; is the time decay factor, is the credible weight coefficient in the high-frequency band.

8. The method according to claim 7, wherein Correct the abnormal duration based on the operation state reliability coefficient, including: Establish a reliability-time correction mapping function: ； Wherein: is the corrected abnormal duration; is the initially set abnormal duration; is the reference credibility threshold; is the correction intensity coefficient; Obtain the corrected abnormal duration based on the operation state reliability coefficient; Optimize the corrected anomaly duration using the exponential smoothing algorithm: ; where: is the optimized anomaly duration, is the smoothing factor; is the anomaly duration after the previous correction.

9. A pressure sensor system, characterized in that, It includes: A pressure sensing unit; An environmental noise monitoring unit; A fault detection module that performs fault detection using the fault detection method according to any one of claims 1 - 8; An alarm output interface.

Images