A method for correcting abnormal readings of a pressure sensor

Abstract

The present invention discloses a method for correcting abnormal reading data of a pressure sensor, which relates to the technical field of sensor equipment, and includes the following steps: collecting sensor readings, building a multi-sensor cross-validation mechanism, establishing an abnormal detection threshold model, abnormal detection based on a time series pattern, building a physical constraint feedback mechanism, performing dynamic drift compensation, processing the collected readings, and self-correction feedback optimization. The method for correcting abnormal reading data of a pressure sensor compares and verifies the sensor output with a theoretical physical model through a physical constraint feedback mechanism, thereby enhancing the ability to identify nonlinear anomalies; through dynamic drift compensation, the linear and exponential drifts generated by the sensor during long-term use are separated and corrected, thereby achieving accurate correction of the drift. The method can significantly improve the reliability and stability of pressure sensor data, provide a more accurate measurement basis for precision measurement scenarios, and reduce the risk of system false alarms and failures.

Description

Technical Field

[0001] The present invention relates to the technical field of sensor devices, and specifically to a method for correcting abnormal pressure sensor readings and data. Background Art

[0002] Pressure sensors are widely used in industries, aviation, medical care, environmental monitoring and other fields. The accuracy and stability of their data directly affect the safety and performance of the system. However, the measured values of pressure sensors are often interfered by various factors, including the drift of the sensor itself (such as aging and mechanical fatigue), and the non-linear effects in complex physical environments. These problems may lead to abnormal sensor readings, thereby affecting the credibility of the data and the overall performance of the system. To solve these problems, traditional methods usually adopt simple statistical methods or conventional data cleaning and filtering techniques. However, they often have the following deficiencies when dealing with complex dynamic environments:

[0003] Lack of physical constraint verification: Traditional methods focus more on statistical laws and ignore the physical characteristics of the system where the sensor is located, resulting in insufficient ability to identify non-linear anomalies.

[0004] Unable to effectively cope with long-term drift: The drift problem caused by aging is often ignored or simply processed, lacking a targeted dynamic compensation mechanism, and unable to guarantee the measurement accuracy during long-term operation. Summary of the Invention

[0005] Aiming at the deficiencies of the prior art, the present invention provides a method for correcting abnormal pressure sensor readings and data to solve the problems raised in the above background art.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides a method for correcting abnormal pressure sensor readings and data, including the following steps:

[0008] S1. Collect sensor readings and perform noise isolation and signal filtering;

[0009] S2. Construct a multi-sensor cross-verification mechanism to initially identify abnormal readings;

[0010] S3. Establish an abnormal detection threshold model, dynamically adjust the threshold range, and identify non-linear anomalies;

[0011] S4. Construct abnormal detection based on time series patterns to identify long-term drift anomalies;

[0012] S5. Construct a physical constraint feedback mechanism, dynamically adjust the tolerance threshold according to the fluid conditions monitored by the sensor, and correct non-linear anomalies;

[0013] S6. Perform dynamic drift compensation to handle long-term drift anomalies;

[0014] S7. Based on the above process, collect sensor readings and process them;

[0015] S8. Perform self-calibration feedback optimization to adapt to new anomaly patterns.

[0016] To further optimize this technical solution, in step S1, wavelet transform decomposition or adaptive filtering algorithm is adopted to separate high-frequency noise and low-frequency interference in the sensor signal;

[0017] And perform spectrum analysis on the signal to identify and remove noise interference at specific frequencies while retaining important signal features.

[0018] To further optimize this technical solution, in step S2, the multi-sensor cross-validation mechanism includes:

[0019] Use multiple pressure sensors to measure the target pressure simultaneously, and identify outliers through data cross-validation between sensors;

[0020] If the reading of a certain sensor deviates significantly from other sensors, it is initially determined that the reading of this sensor may be abnormal;

[0021] Based on the weighted average strategy, assign weights to each sensor according to its historical stability. The higher the weight, the more reliable the sensor data.

[0022] To further optimize this technical solution, in step S3, the anomaly detection threshold model is as follows:

[0023] ;

[0024] Where,

[0025] : The historical mean of the pressure sensor, reflecting the expected value under normal working conditions;

[0026] : The standard deviation of the pressure reading, measuring the fluctuation range and dynamically adjusting the width of the threshold;

[0027] : The sensitivity coefficient of anomaly detection, with a value range of [2.5 - 4.0], used to control the tightness of the detection boundary;

[0028] : The environmental perturbation function, indicating the influence of environmental factors on the sensor reading;

[0029] : The historical drift function, used to correct the possible systematic deviation of the pressure sensor during long-term use.

[0030] To further optimize this technical solution, in step S4, the anomaly detection includes:

[0031] Using a sliding window algorithm or a long short-term memory network (LSTM) to analyze the time series characteristics of sensor data;

[0032] Identifying sudden anomalies or long-term drift phenomena based on the time series characteristics;

[0033] Identifying long-term drift anomalies by detecting the long-term gradual change trend of data fluctuations.

[0034] To further optimize this technical solution, in step S5, the physical constraint feedback mechanism includes:

[0035] When monitoring the fluid pressure in the pipeline, calculating the theoretical value by combining Bernoulli's equation and the continuity equation and comparing it with the sensor output;

[0036] If the deviation exceeds the reasonable range, it is marked as an anomaly and a correction algorithm is fed back.

[0037] To further optimize this technical solution, in the physical constraint feedback mechanism, a feedback is constructed based on the difference between the theoretical pressure value calculated by combining Bernoulli's equation and the continuity equation and the actual measured value:

[0038] ;

[0039] Wherein, is the pressure deviation, is the pressure value output by the sensor, is the theoretical pressure value;

[0040] If , is the set tolerance threshold, it is marked as an anomaly, and the feedback correction mechanism corrects the measured value:

[0041] ;

[0042] Wherein, is the corrected pressure value, is the correction coefficient, which is used to avoid error amplification caused by overcorrection and is usually taken as .

[0043] To further optimize this technical solution, in step S6, based on the drift problem existing in the long-term use of the sensor, a dynamic drift compensation model is constructed, and the dynamic drift compensation model is as follows:

[0044] ;

[0045] Wherein,

[0046] : The corrected pressure value, i.e., the target output value;

[0047] : The current measured value of the sensor;

[0048] : The initial offset of the drift, zero - point offset calibration based on historical initial data;

[0049] : The linear component of the drift, representing the drift amount that increases linearly with time, caused by sensor aging or mechanical strain;

[0050] : The exponential decay component of the drift, used to describe the rapid drift phenomenon of the sensor in the early stage;

[0051] : The running time of the sensor, starting from the startup time.

[0052] Further optimizing this technical solution, in step S7, the processing of the readings includes:

[0053] Data reconstruction to fill abnormal points;

[0054] Data hierarchical screening;

[0055] Data multi - level verification.

[0056] Further optimizing this technical solution, in step S8, the self - calibration feedback optimization includes:

[0057] By real - time collecting the feedback of the sensor and the external system, including reference data provided by manual calibration or other measuring devices;

[0058] Continuously adjusting parameters to ensure the accuracy of the calibration result, adopting a cyclic optimization strategy, regularly self - checking and updating the algorithm to adapt to new abnormal patterns;

[0059] The adjusted parameters include the sensitivity coefficient of the anomaly detection model , and the tolerance threshold of the physical constraint feedback mechanism.

[0060] In a second aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program instructions are executed by the processor, the steps of a method for correcting abnormal pressure sensor reading data as described in the first aspect of the present invention are implemented.

[0061] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, wherein: when the computer program instructions are executed by a processor, the steps of a method for correcting abnormal pressure sensor readings according to the first aspect of the present invention are implemented.

[0062] Compared with the prior art, the present invention provides a method for correcting abnormal pressure sensor readings, having the following beneficial effects:

[0063] The method for correcting abnormal pressure sensor readings overcomes the deficiencies of the prior art through a physical constraint feedback mechanism and dynamic drift compensation, effectively improving the accuracy of anomaly detection in complex dynamic environments; through the physical constraint feedback mechanism, the sensor output is compared and verified with the theoretical physical model, enhancing the ability to identify non-linear anomalies; through dynamic drift compensation, the linear and exponential drifts generated during the long-term use of the sensor are separated and corrected, achieving precise correction of the drift. This method can significantly improve the reliability and stability of pressure sensor data, provide a more accurate measurement basis for various precision measurement scenarios, and at the same time reduce the risk of system false alarms and failures. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0065] Figure 1 It is a schematic flow chart of a method for correcting abnormal pressure sensor readings proposed by the present invention;

[0066] Figure 2 It is a schematic flow chart of an anomaly detection threshold model in a method for correcting abnormal pressure sensor readings proposed by the present invention;

[0067] Figure 3 It is a schematic flow chart of a physical constraint feedback mechanism in a method for correcting abnormal pressure sensor readings proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0068] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the specific embodiments of the present invention in detail with reference to the drawings in the specification.

[0069] In the following description, many specific details are set forth in order to provide a thorough understanding of the present invention. However, the present invention may be practiced in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0070] Secondly, as used herein, an "embodiment" or "embodiments" refers to specific features, structures, or characteristics that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it an embodiment that is separate or mutually exclusive of other embodiments.

[0071] Embodiment 1:

[0072] Referring to Figure 1 , which is the first embodiment of the present invention, this embodiment provides a method for correcting abnormal pressure sensor readings, including the following steps:

[0073] S1. Collect sensor readings and perform noise isolation and signal filtering

[0074] In this embodiment, wavelet transform decomposition or an adaptive filtering algorithm is used to separate high-frequency noise and low-frequency interference in the sensor signal;

[0075] And perform spectral analysis on the signal to identify and remove noise interference at specific frequencies while retaining important signal features.

[0076] This method can maintain data stability in extreme environments.

[0077] S2. Construct a multi-sensor cross-validation mechanism

[0078] In this embodiment, the multi-sensor cross-validation mechanism includes:

[0079] Use multiple pressure sensors to simultaneously measure the target pressure, and identify outliers through data cross-validation between sensors;

[0080] If the reading of a certain sensor significantly deviates from that of other sensors, it is initially determined that the reading of this sensor may be abnormal;

[0081] Based on a weighted average strategy, weights are assigned to each sensor according to its historical stability. The higher the weight, the more reliable the sensor data.

[0082] S3. Establish an outlier detection threshold model

[0083] The model needs to consider the impact of environmental disturbances (temperature, vibration, electromagnetic interference, etc.) that the pressure sensor may be subject to on the reading, and adjust the dynamics of the threshold range by analyzing the correlation between these disturbances and the pressure reading.

[0084] S4. Anomaly detection based on time series patterns

[0085] In this embodiment, anomaly detection includes:

[0086] Use sliding window algorithm or long short-term memory network LSTM to analyze the time series characteristics of sensor data;

[0087] Identify sudden anomalies or long-term drift phenomena based on time series characteristics;

[0088] Long-term drift anomalies are identified by detecting the gradual trend of data fluctuations over a long period of time (i.e., the accumulation of changes in the second-order derivative).

[0089] S5. Constructing physical constraint feedback mechanism

[0090] In this embodiment, the physical constraint feedback mechanism includes:

[0091] When monitoring the fluid pressure in the pipeline, the Bernoulli equation and the continuity equation are combined to calculate the theoretical value and compare it with the sensor output;

[0092] If the deviation exceeds the reasonable range, it will be marked as abnormal and fed back to the correction algorithm.

[0093] S6, dynamic drift compensation

[0094] In this embodiment, based on the drift problem existing in the long-term use of the sensor, a dynamic drift compensation model is constructed, and the dynamic drift compensation model is as follows:

[0095] ;

[0096] Among them,

[0097] : Corrected pressure value, i.e. target output value;

[0098] : Current measurement value of the sensor;

[0099] : Drift initial offset, zero offset calibration based on historical initial data;

[0100] : The linear component of drift, which represents the amount of drift that increases linearly with time, caused by sensor aging or mechanical strain;

[0101] : The exponentially decaying component of drift, which is used to describe the rapid drift phenomenon of the sensor in the early stage;

[0102] : The running time of the sensor, starting from when it is powered on.

[0103] When this model is applied to the pipeline pressure monitoring sensor,

[0104] It is calculated during initial calibration .

[0105] After running for 1000 seconds, the real-time measured value is .

[0106] Drift compensation amount:

[0107]

[0108] Calibrated pressure:

[0109]

[0110] The final output calibrated value is 99.0 Pa, compensating for the drift error.

[0111] Through this model, the pressure sensor can achieve real-time drift compensation, ensuring the accuracy and reliability of data during long-term use, and effectively reducing the false alarm risk caused by drift.

[0112] S7. Collect sensor readings and process them

[0113] In this embodiment, the processing of the readings includes:

[0114] Data reconstruction to fill abnormal points

[0115] Use data reconstruction algorithms to fill abnormal points. For example, through interpolation methods (linear, Lagrange, polynomial interpolation) or prediction models based on historical data (ARIMA model or deep learning prediction model) to reconstruct possible real readings. To improve accuracy, the results of multiple reconstruction algorithms can be weighted and fused to finally obtain reconstructed data with high credibility.

[0116] Data stratification and screening

[0117] Divide the data into high-confidence area, middle area, and low-confidence area according to credibility.

[0118] Data multi-level verification

[0119] Data in the high-confidence area can be directly used for downstream analysis, data in the middle area needs further evaluation, and data in the low-confidence area needs to be processed by reconstruction or elimination. Through this mechanism, the interference of abnormal data to the overall analysis can be reduced, and the calibration efficiency can be improved.

[0120] S8, Self-Calibration Feedback Optimization

[0121] In this embodiment, the self-calibration feedback optimization includes:

[0122] By collecting real-time feedback from sensors and external systems, including reference data provided by manual calibration or other measuring devices;

[0123] Continuously adjust parameters to ensure the accuracy of the calibration results. Adopt a cyclic optimization strategy, regularly self-check and update the algorithm to adapt to new abnormal patterns;

[0124] The adjusted parameters include the sensitivity coefficient of the anomaly detection model , and the tolerance threshold of the physical constraint feedback mechanism.

[0125] By adjusting the sensitivity coefficient , when the environmental conditions change, the parameter adjustment can determine the threshold range of anomaly detection. For example, in an environment with large temperature changes, increasing can increase the threshold width to avoid false alarms.

[0126] By adjusting the tolerance threshold of the physical constraint feedback mechanism, during the anomaly detection process, dynamically adjust the tolerance threshold according to the feedback error, which can help the model adapt to the fault tolerance range under different working conditions and make the anomaly detection mechanism more flexible.

[0127] Embodiment 2:

[0128] Referring to Figures 2 to 3 , this is the second embodiment of the present invention, which provides the anomaly detection threshold model and usage process in step S3.

[0129] In this embodiment, the anomaly detection threshold model is as follows:

[0130] ;

[0131] Wherein,

[0132] : The historical mean of the pressure sensor, reflecting the expected value under normal working conditions;

[0133] : The standard deviation of the pressure readings, measuring the fluctuation range and dynamically adjusting the width of the threshold;

[0134] : The sensitivity coefficient of anomaly detection, with a value range of [2.5 - 4.0], used to control the tightness of the detection boundary;

[0135] : The environmental perturbation function, which represents the influence of environmental factors on the sensor readings, has the following specific form:

[0136]

[0137] In the formula,

[0138] : The influence of environmental temperature on pressure (linear or non-linear, determined based on experimental data).

[0139] Measure the output change of the sensor at different temperatures through experiments. The sensor can be placed in multiple temperature environments, record the deviation between its output and the true pressure, and analyze the influence of temperature change on the readings. Use regression analysis to determine the relationship between temperature and pressure readings.

[0140] : The influence of the intensity of vibration on the signal (determine the correlation coefficient through spectrum analysis).

[0141] Through spectrum analysis, analyze the influence of vibrations at different frequencies on the sensor output signal. First, use an acceleration sensor or a vibration table to conduct vibration tests on the sensor and collect its output data at different frequencies. Through spectrum analysis (such as the Fast Fourier Transform FFT), determine the correlation between the vibration frequency and the sensor output.

[0142] : The influence of the electromagnetic interference field strength (determined through the calibration experiment of the sensor's anti-interference ability).

[0143] Through the calibration experiment of the electromagnetic anti-interference ability, record the change of the sensor output at different electromagnetic field strengths, and calculate the correlation coefficient between the electromagnetic interference intensity and the sensor readings.

[0144] Coefficient Determined through regression analysis or optimization algorithms.

[0145] : The historical drift function, which is used to correct the possible systematic deviation of the pressure sensor during long-term use, is expressed as:

[0146]

[0147] In the formula,

[0148] : The usage time of the sensor.

[0149] : The exponential factor, which describes the non-linear change law of drift and is adjusted based on historical test data.

[0150] : The correction amplitude, which is related to the type and working conditions of the sensor.

[0151] When in use, the model includes:

[0152] Calculation of historical mean and standard deviation: Using the historical operation data of the pressure sensor, calculate the mean and standard deviation according to the time window. For example, select the data of the past 24 hours and update it in real time to obtain and . This can track the central tendency and fluctuation range of the pressure under normal working conditions.

[0153] Calculation of disturbance factors: Real-time monitor environmental parameters (such as temperature, vibration, electromagnetic interference), and input these parameters into the disturbance function . Through the real-time feedback of the sensor operating environment, dynamically adjust the anomaly detection threshold. For example, in a high-vibration environment, the impact of vibration on the readings may increase significantly, expanding the detection range and reducing false alarms.

[0154] Drift compensation: Use the historical operating time and cumulative drift data of the sensor to determine the parameters of the drift function through non-linear regression (such as and ). For sensors operating for a long time, their pressure readings may continuously deviate. Systematic errors should be compensated by subtracting . For example, after operating for 1000 hours, if the detected offset is 0.5 MPa, the compensation amount is deducted from the threshold.

[0155] Calculation of dynamic threshold: Substitute all components into the formula to calculate the upper and lower threshold values in real time and make a judgment on anomalies.

[0156] Adaptive adjustment: By introducing a feedback mechanism, adjust in real time according to the accuracy of anomaly detection (such as false alarm rate, miss rate). For example, when the environmental disturbance is large and there are frequent false alarms in detection, the system can automatically increase

[0157] Design and apply the physical constraint feedback mechanism in step S5.

[0158] In the physical constraint feedback mechanism, the theoretical pressure value formula calculated by combining the Bernoulli equation and the continuity equation is as follows:

[0159] ;

[0160] where is the pressure value, is the fluid density, is the flow velocity, is the acceleration due to gravity, is the fluid height;

[0161] Construct a feedback based on the difference between the theoretical pressure value and the actual measured value:

[0162] ;

[0163] If , is the set tolerance threshold, it is marked as abnormal, and the feedback correction mechanism corrects the measured value:

[0164] ;

[0165] Among them, is the correction coefficient, which is used to avoid the error amplification caused by overcorrection, and usually takes .

[0166] When this mechanism is used, it includes:

[0167] Input physical parameters and initial conditions: Collect relevant physical parameters.

[0168] Calculate the theoretical pressure value in real time: Use the real-time flow rate and position height data of the sensor, and calculate the theoretical pressure of the target pressure point through the Bernoulli equation and the continuity equation . If the installation position of the sensor is fixed, static physical parameters can be pre-stored to improve the calculation efficiency.

[0169] Compare the actual pressure with the theoretical value: Compare the pressure value output by the sensor with to calculate the pressure deviation . If the deviation is within the tolerance range, the measured value is considered reasonable; otherwise, it is marked as abnormal.

[0170] Abnormal feedback correction: Correct the abnormal data points and adjust the deviation amount proportionally . For abnormal conditions that cannot be corrected (such as a completely failed sensor), trigger the alarm mechanism and exclude the output data of this sensor.

[0171] Multi-point data smoothing and verification: If the system consists of multiple sensors (such as installed at multiple positions along the pipeline), further optimize the correction result through cross-sensor data comparison and fluid continuity verification. For example, use the theoretical value of the downstream sensor as the verification basis to reduce the error impact of isolated sensors.

[0172] Dynamically adjust the tolerance threshold: Dynamically adjust the tolerance threshold according to the fluid working conditions (such as drastic changes in flow rate or external interference) . For example, when the flow rate change rate ( ) is large, appropriately relax the threshold to avoid false alarms.

[0173] Example 3:

[0174] This embodiment also provides a computer device, which is applicable to a method for correcting abnormal pressure sensor readings. It includes a memory and a processor. The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement a method for correcting abnormal pressure sensor readings as proposed in the above embodiment.

[0175] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements a method for correcting abnormal pressure sensor readings as proposed in the above embodiment.

[0176] The computer device can be a terminal. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the housing of the computer device, or an external keyboard, a touchpad, or a mouse, etc.

[0177] If a function 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 such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. The 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 in various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs, etc., which can store program codes.

[0178] The logic and / or steps represented in the flowchart or otherwise described herein can, for example, be considered as a definitional sequence of executable instructions for implementing logical functions, which can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with the instruction execution system, apparatus, or device.

[0179] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection (electronic device) having one or more wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which a program can be printed, as the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise appropriate processing as necessary, and then stored in a computer memory.

[0180] It should be understood that the various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the multiple steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one of the following techniques known in the art or a combination thereof can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), and the like.

[0181] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A method for correcting abnormal pressure sensor readings, characterized in that: The following steps are involved: S1, collect sensor readings, and perform noise isolation and signal filtering; S2, build a multi-sensor cross-validation mechanism to preliminarily identify abnormal readings; S3, establish anomaly detection threshold model, dynamically adjust the threshold range, and identify nonlinear anomalies; S4, build anomaly detection based on time series patterns to identify long-term drift anomalies; S5. Build a physical constraint feedback mechanism to dynamically adjust the tolerance threshold according to the fluid conditions monitored by the sensor and correct nonlinear anomalies; Physical constraint feedback mechanisms include: When monitoring the fluid pressure in a pipeline, the Bernoulli equation and the continuity equation are combined to calculate the theoretical value and compare it with the sensor output; If the deviation exceeds a reasonable range, it is marked as abnormal and fed back to the correction algorithm; The formula for calculating the theoretical pressure value by combining the Bernoulli equation and the continuity equation is as follows: ； in, is the pressure value, is the fluid density, is the flow rate, is the acceleration due to gravity, is the fluid height; S6. Perform dynamic drift compensation to handle long-term drift anomalies; S7, based on the above process, collect sensor readings and process them; S8. Perform self-correction feedback optimization to adapt to new abnormal patterns.

2. A method for correcting abnormal pressure sensor readings according to claim 1, characterized in that: In step S1, wavelet transform decomposition or adaptive filtering algorithm is used to separate high-frequency noise and low-frequency interference in the sensor signal; The signal is then spectrally analyzed to identify and remove noise interference at specific frequencies while retaining important signal features.

3. A method for correcting abnormal pressure sensor readings according to claim 1, characterized in that: In step S2, the multi-sensor cross-validation mechanism includes: Use multiple pressure sensors to measure the target pressure simultaneously and identify outliers by cross-validating data between sensors; The reading of a certain sensor deviates significantly from the other sensors, and it is preliminarily determined that the reading of the sensor may be abnormal; Based on the weighted average strategy, weights are assigned to each sensor according to its historical stability. The higher the weight, the more credible the sensor data.

4. A method for correcting abnormal pressure sensor readings according to claim 1, characterized in that: In step S3, the anomaly detection threshold model is as follows: ； in, : The historical average value of the pressure sensor, reflecting the expected value under normal working conditions; : The standard deviation of the pressure readings, which measures the fluctuation range and dynamically adjusts the width of the threshold; : The sensitivity coefficient of anomaly detection, ranging from [2.5-4.0], is used to control the tightness of the detection boundary; : Environmental disturbance function, which represents the impact of environmental factors on sensor readings; : Historical drift function, used to correct possible systematic deviations of the pressure sensor during long-term use.

5. The method for correcting abnormal pressure sensor readings according to claim 1, characterized in that: In step S4, abnormality detection includes: Use sliding window algorithm or long short-term memory network LSTM to analyze the time series characteristics of sensor data; Identify sudden anomalies or long-term drift phenomena based on time series characteristics; Identify long-term drift anomalies by detecting gradual trends in data fluctuations over time.

6. A method for correcting abnormal pressure sensor readings according to claim 1, characterized in that: In the physical constraint feedback mechanism, feedback is constructed by combining the difference between the theoretical pressure value calculated by the Bernoulli equation and the continuity equation and the actual measured value: ； in, is the pressure deviation, is the pressure value output by the sensor, is the theoretical pressure value; like , If the tolerance threshold is exceeded, it is marked as abnormal, and the feedback correction mechanism corrects the measured value: ； in, To calibrate the pressure value, is the correction factor, which is used to avoid error amplification caused by over-correction. .

7. A method for correcting abnormal pressure sensor readings according to claim 1, characterized in that: In step S6, based on the drift problem existing in the long-term use of the sensor, a dynamic drift compensation model is constructed, and the dynamic drift compensation model is as follows: ； in, : Corrected pressure value, i.e. target output value; : Current measurement value of the sensor; : Drift initial offset, zero offset calibration based on historical initial data; : The linear component of drift, which represents the amount of drift that increases linearly with time, caused by sensor aging or mechanical strain; : The exponential decay component of the drift, which is used to describe the rapid drift phenomenon of the sensor in the early stage; : Sensor operation time, starting from startup.

8. The method for correcting abnormal pressure sensor readings according to claim 1, characterized in that: In step S7, the processing of the readings includes: Data reconstruction fills in abnormal points; Data stratification and screening; Multi-level data verification.

9. A method for correcting abnormal pressure sensor readings according to claim 1, characterized in that: In step S8, the self-correction feedback optimization includes: By collecting sensor and external system feedback in real time, including reference data provided by manual calibration or other measurement equipment; Continuously adjust parameters to ensure the accuracy of calibration results, adopt cyclic optimization strategy, regularly self-check and update algorithms to adapt to new abnormal patterns; The parameters adjusted include the sensitivity coefficient of the anomaly detection model , and the tolerance threshold of the physical constraint feedback mechanism.

Images