An online maintenance method and system based on pressure transmitter

Abstract

The application relates to the technical field of industrial process measurement, and particularly discloses a pressure transmitter online maintenance method and system, a pressure signal sampling window and a background noise monitoring window are synchronously started, pressure data and auxiliary diagnosis parameters are respectively collected; physical validity inspection is carried out on the pressure data based on signal continuity, abnormal data segments are removed, a reliable data set is generated, and then a robust center estimation value is calculated; a data quality acceptance threshold is dynamically adjusted according to the auxiliary diagnosis parameters, and the robust center estimation value and data distribution are compoundly checked based on the threshold; only when the checking is passed, zero point writing with verification is executed and complete traceability records are generated, otherwise, the process is interrupted and a diagnostic warning is output; through multi-dimensional data fusion and intelligent decision, the application effectively filters out transient interference, ensures the absolute reliability of the online calibration benchmark, and improves the traceability of the maintenance process.

Description

Technical Field

[0001] This invention relates to the field of industrial process measurement technology, specifically to a method and system for online maintenance of pressure transmitters. Background Technology

[0002] In continuous production processes in process industries, pressure transmitters serve as critical process monitoring and control instruments. Their accuracy and reliability directly impact production safety, product quality, and operational efficiency. To maintain measurement accuracy, regular zero-point calibration is an essential maintenance procedure. Traditional offline calibration methods require removing the transmitter from the process pipeline and sending it to a laboratory or on-site for calibration using a standard. This process can lead to production interruptions and poses a risk of media leakage, making it both economically and safety-wise inadequate.

[0003] Existing online zero-point calibration methods for pressure transmitters rely on the simple trust and direct setting of a single instantaneous sample value under the "physical zero differential pressure" state. In complex industrial environments such as those with strong electromagnetic interference, these methods are prone to misinterpreting random pulse interference values ​​at the moment of sampling as the true zero point and fixing them. This results in the pressure transmitter outputting measurement signals with hidden systematic errors during subsequent long-term operation, threatening the accuracy of process control and the reliability of safety instrument systems. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for online maintenance of pressure transmitters to solve the problems mentioned above.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A method for online maintenance of pressure transmitters includes the following steps:

[0007] S1: Perform zero-point calibration on the pressure transmitter to be inspected, and simultaneously start a sampling window of a predetermined duration and a background noise monitoring window;

[0008] S2: Within the sampling window, continuously acquire the raw pressure signal value of the pressure transmitter at the first sampling frequency to form the first dataset; within the background noise monitoring window, monitor and record the auxiliary diagnostic parameters related to signal integrity inside the pressure transmitter.

[0009] S3: Perform a physical validity test on the first dataset based on the continuity of the pressure signal, remove abnormal data segments that do not conform to the physical change law, generate the second dataset, and calculate the robust center estimate of the second dataset.

[0010] S4: Based on the status of the auxiliary diagnostic parameters, adjust the acceptable fluctuation threshold used to assess the quality of the second dataset, and perform a composite verification of the robust center estimate and the statistical distribution of the second dataset based on the adjusted acceptable fluctuation threshold.

[0011] S5: Only when the composite verification passes, the robust center estimate is written to the zero-point storage unit of the pressure transmitter to complete the calibration, and a calibration traceability record containing auxiliary diagnostic parameters, acceptable fluctuation thresholds and verification results is generated; if the verification fails, the calibration process is interrupted and a fault warning containing auxiliary diagnostic parameters is output.

[0012] As a further aspect of the present invention: the formation of the first dataset specifically includes:

[0013] The sampling window is divided into multiple consecutive and partially overlapping data frames;

[0014] Within each data frame, the acquired raw pressure signal value is processed in real time using a sliding median filter to obtain the corresponding representative value within the frame;

[0015] Collect the intra-frame representative value sequence of all data frames and calculate the mean absolute difference of the intra-frame representative value sequence;

[0016] If the mean absolute difference is less than the preset first threshold, all intra-frame representative values ​​are stored as the first dataset; if the mean absolute difference is greater than or equal to the first threshold, the current data is discarded and the sampling window is automatically extended for a predetermined duration. The acquisition and division steps are then repeated until the obtained intra-frame representative value sequence satisfies the condition that the mean absolute difference is less than the first threshold.

[0017] As a further aspect of the present invention: the monitoring and recording of auxiliary diagnostic parameters related to signal integrity within the pressure transmitter specifically includes:

[0018] Within the background noise monitoring window, the voltage fluctuation signal at the reference voltage terminal of the analog-to-digital converter of the pressure transmitter and the reading change signal of the internal temperature sensor of the pressure transmitter within the monitoring window are simultaneously acquired.

[0019] The effective value of the voltage fluctuation signal and the average rate of change of the reading change signal are calculated to form auxiliary diagnostic parameters with two dimensions.

[0020] As a further aspect of the present invention: S3 specifically includes:

[0021] Calculate the rate of pressure change between adjacent data points in the first dataset to form a rate of change sequence;

[0022] Identify abrupt changes in the rate of change sequence where the absolute value exceeds a preset physical threshold, and mark each abrupt change and a predetermined number of data points before and after it as the segment to be inspected;

[0023] Check whether the pressure value of each data point in the section to be inspected exhibits an oscillation or step pattern that does not conform to the law of inertia;

[0024] All data from the inspected sections that were determined to conform to the laws of physical change were re-integrated to form a continuous and physically reliable second dataset.

[0025] As a further aspect of the present invention: the step of checking whether the pressure value of each data point in the inspected section exhibits an oscillation or step pattern that does not conform to the law of inertia specifically includes:

[0026] Calculate the absolute value sequence of the first difference of pressure data points within the section to be inspected;

[0027] Identify all local maxima in the absolute value sequence, calculate the average amplitude of these local maxima, and determine whether there are three or more consecutive local maxima whose amplitudes all exceed a set multiple of the average amplitude.

[0028] If it exists, the corresponding segment is determined to exhibit an oscillation pattern that does not conform to the law of inertia;

[0029] If it does not exist, further check whether the pressure difference between the start and end points of the corresponding section is greater than another set multiple of the average amplitude and whether the direction of change is contrary to the direction of the physical process, in order to determine whether it is a step form that does not conform to the law.

[0030] As a further aspect of the present invention: the calculation of the robust center estimate of the second dataset specifically includes:

[0031] All pressure data points in the second dataset are sequentially numbered based on time, and all pressure data points are treated as a time-domain signal sequence for frequency domain transformation analysis to obtain the spectral distribution of the time-domain signal sequence;

[0032] Analyze the main frequency components of energy concentration in the spectrum distribution and identify whether they fall within the pre-defined reasonable frequency band of the physical process;

[0033] If the main frequency components fall within a reasonable frequency band, the second dataset is determined to reflect the real physical process, and the arithmetic mean of the second dataset is used as the robust center estimate; if the main frequency components fall outside a reasonable frequency band, the median of the second dataset is used as the robust center estimate.

[0034] As a further aspect of the present invention: S4 specifically includes:

[0035] The interference assessment value, which characterizes the instantaneous electrical interference intensity, is extracted from the auxiliary diagnostic parameters, and the interference assessment value is used to amplify the preset benchmark threshold to obtain the first correction threshold.

[0036] The rate of change characterizing temperature stability is extracted from the auxiliary diagnostic parameters, and the rate of change is used to scale the first correction threshold to generate the final acceptable fluctuation threshold.

[0037] Simultaneously verify whether the robust center estimate is within the range centered on the average of the second dataset and with the acceptable fluctuation threshold as the radius, and whether the range of the second dataset is less than a preset multiple of the acceptable fluctuation threshold. When both conditions are met, the composite verification is passed.

[0038] As a further aspect of the present invention: obtaining the first correction threshold specifically includes:

[0039] Obtain the voltage ripple signal from the auxiliary diagnostic parameters, and calculate the root mean square value of the voltage ripple signal within a predetermined short time window as the original interference intensity;

[0040] The zero-crossing interval sequence of the voltage ripple signal is analyzed, and the coefficient of variation of the zero-crossing interval sequence interval value is statistically analyzed to characterize the suddenness and irregularity of the interference.

[0041] The original interference intensity and the coefficient of variation are weighted and fused to generate the interference assessment value;

[0042] The first correction threshold is obtained by multiplying the interference evaluation value by a preset baseline threshold.

[0043] An online maintenance system for pressure transmitters includes:

[0044] Synchronous trigger module: performs zero-point calibration on the pressure transmitter to be inspected, and simultaneously starts a sampling window of a predetermined duration and a background noise monitoring window;

[0045] Dual-channel parallel acquisition module: Within the sampling window, the original pressure signal value of the pressure transmitter is continuously acquired at the first sampling frequency to form the first dataset; within the background noise monitoring window, auxiliary diagnostic parameters related to signal integrity inside the pressure transmitter are monitored and recorded.

[0046] Physical validity test and robust estimation module: Performs a physical validity test on the first dataset based on the continuity of the pressure signal, removes abnormal data segments that do not conform to the physical change law, generates the second dataset, and calculates the robust center estimate of the second dataset;

[0047] Dynamic threshold composite verification module: Based on the status of auxiliary diagnostic parameters, adjust the acceptable fluctuation threshold used to evaluate the quality of the second dataset, and perform composite verification on the robust center estimate and the statistical distribution of the second dataset based on the adjusted acceptable fluctuation threshold;

[0048] Calibration execution and traceability record module: Only when the composite verification passes, the robust center estimate is written to the zero-point storage unit of the pressure transmitter to complete the calibration, and a calibration traceability record containing auxiliary diagnostic parameters, acceptable fluctuation thresholds and verification results is generated; if the verification fails, the calibration process is interrupted and a fault warning containing auxiliary diagnostic parameters is output.

[0049] The beneficial effects of this invention are:

[0050] (1) By synchronously acquiring pressure signals and background noise, and performing continuity and morphological checks on the raw pressure data based on physical laws, non-physical distortion data introduced by transient pulse interference and signal chain anomalies can be effectively identified and eliminated. Combined with acceptance thresholds dynamically adjusted based on real-time electrical noise and temperature stability, the data quality and central estimate are compounded and verified, thereby ensuring that the final calibration benchmark is based solely on real and stable physical measurements. This reduces the risk of misjudging environmental interference or equipment transient anomalies as zero-point drift and improves the reliability and accuracy of online calibration in complex industrial electromagnetic environments.

[0051] (2) This invention not only outputs the result of whether the pressure transmitter calibration is successful or not, but also generates a structured traceability record containing complete auxiliary diagnostic parameters, dynamic thresholds, verification results, and post-write verification data. When calibration fails, it can provide a clear fault warning, such as high ripple noise or abnormal temperature drift. This provides key data support for maintenance personnel to quickly locate the root cause of the problem (such as on-site electromagnetic environment problems, transmitter internal circuit faults, or installation process defects), transforming traditional "trial" repairs into "diagnostic" maintenance, shortening fault diagnosis time, and improving maintenance efficiency and equipment availability. Attached Figure Description

[0052] The invention will now be further described with reference to the accompanying drawings.

[0053] Figure 1 This is a flowchart of the method of the present invention;

[0054] Figure 2 This is a system block diagram of the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Please see Figure 1As shown, this invention is a method for online maintenance of pressure transmitters, comprising the following steps:

[0057] S1: Perform zero-point calibration on the pressure transmitter to be inspected, and simultaneously start a sampling window of a predetermined duration and a background noise monitoring window;

[0058] S2: Within the sampling window, continuously acquire the raw pressure signal value of the pressure transmitter at the first sampling frequency to form the first dataset; within the background noise monitoring window, monitor and record the auxiliary diagnostic parameters related to signal integrity inside the pressure transmitter.

[0059] S3: Perform a physical validity test on the first dataset based on the continuity of the pressure signal, remove abnormal data segments that do not conform to the physical change law, generate the second dataset, and calculate the robust center estimate of the second dataset.

[0060] S4: Based on the status of the auxiliary diagnostic parameters, adjust the acceptable fluctuation threshold used to assess the quality of the second dataset, and perform a composite verification of the robust center estimate and the statistical distribution of the second dataset based on the adjusted acceptable fluctuation threshold.

[0061] S5: Only when the composite verification passes, the robust center estimate is written to the zero-point storage unit of the pressure transmitter to complete the calibration, and a calibration traceability record containing auxiliary diagnostic parameters, acceptable fluctuation thresholds and verification results is generated; if the verification fails, the calibration process is interrupted and a fault warning containing auxiliary diagnostic parameters is output.

[0062] In step S1, when the pressure transmitter receives a zero-point calibration command transmitted via the digital communication interface while operating online, its internal microprocessor unit responds immediately. This unit simultaneously initiates two independent timing processes: the first process controls a sampling window for acquiring pressure signals, the window duration of which is preset to cover multiple power frequency cycles and periods where the pressure signal may fluctuate; the second process controls a background noise monitoring window for monitoring internal electrical noise, the duration of which is consistent with or proportional to the sampling window. The strict alignment of the start times of these two windows ensures that the subsequently acquired pressure data and the monitored background noise are synchronously comparable on the time axis, providing a time-correlation basis for evaluating signal quality in subsequent steps.

[0063] In S2, the construction of the first dataset employs a verification method based on data frame stability. First, the entire sampling window is divided into multiple consecutive, temporally overlapping sub-intervals, each called a data frame. For a set of raw pressure signal values ​​collected within each data frame, a sliding median filter is applied: the values ​​are sorted from smallest to largest, and the value at the middle position after sorting is taken as the representative value within that data frame. After completing the above processing for all data frames, a sequence composed of the representative values ​​within each frame is obtained. To evaluate the stability of this sequence, its mean absolute difference (MAD) is calculated: first, the arithmetic mean of all representative values ​​within the sequence is calculated; then, the absolute difference between each representative value and the arithmetic mean is calculated; finally, the arithmetic mean of these absolute differences is calculated, and the result is the MAD. A first threshold is preset as a stability criterion. If the calculated MAD is less than the first threshold, the data stability within the current sampling window is deemed to meet the requirements, and the set of all representative values ​​within each frame is stored as the first dataset. If the mean absolute difference is greater than or equal to the first threshold, it indicates that the data fluctuation is too large. All collected data in the current sampling window is automatically discarded, and the duration of the sampling window is increased by a preset extension. Then, the entire data acquisition and frame division process is re-executed until the newly obtained data can meet the condition that the mean absolute difference is less than the first threshold, and the final first dataset is generated accordingly.

[0064] Within the synchronously initiated background noise monitoring window, auxiliary diagnostic parameters are acquired in parallel. This includes simultaneously acquiring two key signals: first, the voltage fluctuation signal on the reference voltage pin of the internal analog-to-digital converter of the pressure transmitter, which is acquired through a high-resolution auxiliary analog-to-digital converter channel; and second, the output reading change signal of the built-in temperature sensor of the pressure transmitter throughout the entire monitoring window. For the acquired reference voltage fluctuation signal, its effective value is calculated. The effective value is calculated by first calculating the arithmetic mean of the squares of all sampling points of the voltage fluctuation signal within a complete monitoring window, and then taking the square root of this mean. The result is the effective value characterizing the voltage noise intensity. For the temperature sensor reading change signal, its average rate of change is calculated. The average rate of change is calculated by taking the difference between the temperature reading at the end and the beginning of the monitoring window, and then dividing it by the total duration of the monitoring window. The calculated effective value of the voltage fluctuation and the average rate of change of the temperature are combined to form an auxiliary diagnostic parameter containing two specific numerical elements, used to comprehensively reflect the electrical noise level and thermal state change rate inside the transmitter at the calibration time.

[0065] In S3, the first step is to perform preliminary localization of anomalous mutations in the first dataset. This first dataset consists of a series of stress values ​​collected sequentially over time. Composition, in which This represents the total number of data points. Calculate the rate of pressure change between adjacent data points in this dataset to form a rate of change sequence. , of which rate of change Through calculation and The difference is obtained by dividing by the sampling time interval. A preset physical threshold is used. This threshold is set based on the maximum possible rate of change of pressure in the measured process. The sequence of rates of change is iterated. Identify all that satisfy: The mutation points. For each identified mutation point (corresponding to the original pressure data point index). ), and itself and the ones before and after it. Data points ( For a predetermined number, such as 5), they are collectively marked as a single inspection segment. If multiple mutation points are too close together, causing their corresponding inspection segments to overlap, these overlapping segments are merged into a larger inspection segment for processing.

[0066] Secondly, a detailed morphological analysis is conducted on each marked section to determine whether it exhibits an abnormal pattern that does not conform to the laws of physical inertia, primarily targeting two typical disturbance patterns: oscillation and step. For a given segment containing... One pressure data point: For the section to be inspected, calculate the absolute value sequence of its first-order difference. ,in, ;(for arrive ). Identification sequence All local maxima in the region. A local maximum. Conditions to be met: and Suppose there are a total of local maxima points identified. The set of amplitudes is as follows: Calculate the average magnitude of these local maxima. The calculation formula is as follows: ; Determine in sequence Does the data contain three or more consecutive local maxima, all with amplitudes greater than 1? With a set multiple The product of the two factors is used. If it exists, the section under inspection is determined to exhibit a rapid oscillation pattern that does not conform to the law of inertia. If it does not meet the oscillation pattern criterion, a step pattern test is further performed. The starting pressure value of the section under inspection is calculated. Compared with the endpoint pressure value The difference Determine the difference. Is the absolute value greater than With another set multiple The product of these factors, and whether its direction of change contradicts the physical law that the pressure should remain stable under isolated equilibrium conditions (e.g., in the ideal case where the system is isolated, balanced, and leak-free). The amplitude should approach zero; any significant, unidirectional step change can be considered suspicious. If both the amplitude and direction inconsistencies are met, the segment is determined to exhibit a step pattern that does not conform to physical laws. All data points within the tested segments judged to exhibit oscillations or step patterns will be removed from the overall data. Finally, all the data from the tested segments that were not removed, as well as those judged to conform to physical change patterns, are merged with other normal data segments unaffected by abrupt changes, in their original chronological order to form a temporally continuous and physically reliable pressure data set, namely the second dataset, denoted as [dataset name missing]. ,in This represents the total number of data points in the second dataset.

[0067] Next, robust center estimates are calculated based on the generated second dataset. First, all pressure data points in the second dataset are sequentially numbered based on acquisition time, forming a discrete-time signal sequence. Frequency domain transformation analysis is then performed on this sequence to obtain its spectral distribution. This analysis can be achieved using the Discrete Fourier Transform. For a dataset consisting of… The second dataset sequence consists of data points. ,in Its discrete Fourier transform result is a series of complex numbers , of which The amplitude corresponding to each frequency component By calculating complex numbers The modulus is obtained, that is: ;

[0068] in, and They are respectively The real and imaginary parts. All This constitutes the spectral distribution of the pressure signal sequence. Analyzing this spectral distribution helps identify the dominant frequency components with the highest energy concentration. This can be done by finding the amplitude... maximum value and its corresponding frequency index To achieve this. A reasonable frequency band for the physical process is preset. This frequency band is set based on the range of fluctuations in the monitored process pressure that may occur under normal stable conditions (such as those caused by pump pulsation or minor oscillations of regulating valves). The actual frequencies corresponding to the main frequency components are: ;(in The first sampling frequency is compared with a reasonable frequency band. If the fluctuations reflected in the second dataset are determined to originate primarily from real, slow physical process disturbances, then the arithmetic mean of the second dataset is used as the robustness center estimate. ,Right now: This indicates that there may be extremely low-frequency drift or residual high-frequency noise in the data. These are not characteristics of a stable physical process. In this case, the median of the second dataset is used as the robust center estimate. The method for calculating the median is: take all values ​​from the second dataset... Arrange in ascending order, if the number of data points If the number is odd, the median is the value that is exactly in the middle after sorting; if... If the number is even, the median is the arithmetic mean of the two middle values ​​after sorting. This spectral analysis-based decision-making mechanism adaptively selects the most suitable central tendency measure for the current data characteristics, thus obtaining a more robust zero-point estimation benchmark.

[0069] In step S4, firstly, an interference assessment value characterizing the instantaneous electrical interference intensity is extracted from the auxiliary diagnostic parameters. This value is then used to amplify a preset baseline threshold to obtain a first correction threshold. Specifically, the auxiliary diagnostic parameters include voltage ripple signals acquired within the background noise monitoring window. A data segment of this voltage ripple signal within a predetermined short time window is extracted. The duration of this short time window is set to an integer multiple of the power frequency cycle, for example, 20 milliseconds. The arithmetic mean of the squares of all voltage samples within this data segment is calculated, and then the square root of this mean is taken. The result is the root mean square value of the voltage ripple within this short time window, which is used as the original interference intensity. Simultaneously, the zero-crossing interval sequence of the voltage ripple signal within the entire monitoring window is analyzed. The zero-crossing interval refers to the time taken for the signal to cross zero level twice consecutively from positive to negative or from negative to positive. The standard deviation of this zero-crossing interval sequence is calculated, and the arithmetic mean of the sequence is also calculated. The ratio of the standard deviation to the arithmetic mean is the coefficient of variation of the zero-crossing interval, used to characterize the suddenness and irregularity of the interference. Next, the original interference intensity and the coefficient of variation are weighted and fused to generate an interference assessment value. The weighted fusion uses a linear weighted summation method, i.e., a first weighting coefficient and a second weighting coefficient are preset. The original interference intensity is multiplied by the first weighting coefficient, and the coefficient of variation is multiplied by the second weighting coefficient. The two products are then added together, and the sum is the interference assessment value. The sum of the first and second weighting coefficients is one. Finally, the interference assessment value is multiplied by a reference threshold preset according to the transmitter's accuracy and range; the resulting product is the first correction threshold.

[0070] Secondly, the rate of change characterizing temperature stability is extracted from the auxiliary diagnostic parameters, and this rate of change is used to scale the first correction threshold to generate the final acceptable fluctuation threshold. The auxiliary diagnostic parameters include the average rate of change of the temperature sensor readings calculated in step S2. The method for scaling the first correction threshold using this average rate of change is as follows: calculate the sum of the average temperature rate of change and the absolute value of the average rate of change, then divide the first correction threshold by this sum; the quotient is the final acceptable fluctuation threshold. When the absolute value of the average temperature rate of change is large, it indicates temperature instability. In this case, the denominator increases, and the acceptable fluctuation threshold decreases accordingly, thereby imposing stricter requirements in data verification.

[0071] Finally, based on the generated final acceptable fluctuation threshold, a composite verification is performed on the robust center estimate and the statistical distribution of the second dataset. The composite verification includes two conditions that must be met simultaneously. The first condition verifies whether the robust center estimate falls within a range centered on the mean of the second dataset and with the acceptable fluctuation threshold as its radius. Specifically, the verification method is as follows: calculate the arithmetic mean of all stress data points in the second dataset, then calculate the absolute value of the difference between the robust center estimate and this arithmetic mean, and determine whether this absolute value is less than or equal to the acceptable fluctuation threshold. The second condition verifies whether the range of the second dataset is less than a preset multiple of the acceptable fluctuation threshold. The range refers to the difference between the maximum and minimum stress values ​​in the second dataset. The preset multiple is set empirically, for example, twice. The range is then determined to be less than the acceptable fluctuation threshold multiplied by this preset multiple. Only when both conditions are met is the composite verification considered successful. If either condition is not met, the composite verification fails.

[0072] In step S5, the microprocessor inside the pressure transmitter first receives the composite verification result from step S4. If the result is satisfactory, the microprocessor initiates the zero-point calibration write procedure. This procedure does not directly write the robust center estimate calculated in step S3 into the permanent zero-point storage unit, but first writes it into a temporary zero-point register. After this write operation is completed, the control logic immediately initiates a brief verification sampling: the transmitter's analog-to-digital converter acquires a set of pressure signal values ​​at a first sampling frequency or another preset verification frequency within a predetermined very short time (e.g., 50 milliseconds), and calculates the arithmetic mean of these verification sample values. Subsequently, the average value of the verification samples is compared with the value stored in the temporary zero-point register, and the absolute value of the difference between the two is calculated. If the absolute value is not greater than a preset post-write verification tolerance (which is much smaller than the acceptable fluctuation threshold), the post-write verification is deemed successful. At this time, the microprocessor finally solidifies the value in the temporary zero-point register into the non-volatile zero-point storage unit, formally completing the zero-point reference update. If the post-write verification fails, the zero-point storage unit is not updated, and a post-write verification failure event is recorded.

[0073] After successfully updating the zero-point storage unit, the microprocessor automatically generates a calibration traceability record. This record is stored as a structured data block in a designated area of ​​the transmitter's internal non-volatile memory. The record includes: the precise timestamp of this calibration operation; all auxiliary diagnostic parameters acquired and processed in step S2 (such as the RMS value of voltage ripple and the average rate of temperature change); the final acceptable fluctuation threshold determined in step S4; detailed results of the composite verification (pass / fail); the robustness center estimate calculated in step S3; the average value of the verification samples collected during the post-write verification process; and the post-write verification tolerance used. These data are stored in association within the record using predefined field formats to ensure information integrity and traceability.

[0074] If the composite verification result in step S4 fails, the microprocessor will immediately interrupt the entire calibration process and will not perform any write operations to the zero-point storage unit. Simultaneously, it will generate and output a fault warning message. This message is sent to the connected maintenance handheld device or host computer system via the transmitter's digital communication interface (such as the HART protocol). The warning message not only includes a status indication of calibration failure but also specifically lists the possible related diagnostic parameters and their values ​​that led to the verification failure, such as the currently monitored high-voltage ripple interference intensity or abnormal temperature change rate, providing maintenance personnel with clear clues to locate on-site environmental problems or potential transmitter faults.

[0075] Please see Figure 2 As shown, an online maintenance system based on a pressure transmitter includes:

[0076] Synchronous trigger module: performs zero-point calibration on the pressure transmitter to be inspected, and simultaneously starts a sampling window of a predetermined duration and a background noise monitoring window;

[0077] Dual-channel parallel acquisition module: Within the sampling window, the original pressure signal value of the pressure transmitter is continuously acquired at the first sampling frequency to form the first dataset; within the background noise monitoring window, auxiliary diagnostic parameters related to signal integrity inside the pressure transmitter are monitored and recorded.

[0078] Physical validity test and robust estimation module: Performs a physical validity test on the first dataset based on the continuity of the pressure signal, removes abnormal data segments that do not conform to the physical change law, generates the second dataset, and calculates the robust center estimate of the second dataset;

[0079] Dynamic threshold composite verification module: Based on the status of auxiliary diagnostic parameters, adjust the acceptable fluctuation threshold used to evaluate the quality of the second dataset, and perform composite verification on the robust center estimate and the statistical distribution of the second dataset based on the adjusted acceptable fluctuation threshold;

[0080] Calibration execution and traceability record module: Only when the composite verification passes, the robust center estimate is written to the zero-point storage unit of the pressure transmitter to complete the calibration, and a calibration traceability record containing auxiliary diagnostic parameters, acceptable fluctuation thresholds and verification results is generated; if the verification fails, the calibration process is interrupted and a fault warning containing auxiliary diagnostic parameters is output.

[0081] The working principle of this invention is as follows: Upon receiving an online zero-point calibration command, a pressure signal sampling window and a background noise monitoring window are simultaneously activated. Within the sampling window, the original pressure signal is acquired through high-frequency sampling and processed based on data frame stability to form a first dataset. Simultaneously, within the monitoring window, voltage ripple and temperature change signals inside the transmitter are acquired to form auxiliary diagnostic parameters. The first dataset undergoes a physical validity test based on signal continuity. By identifying and eliminating abnormal data segments that do not conform to the inertial law of oscillation or step patterns, a physically reliable second dataset is generated. Based on its spectral characteristics, the arithmetic mean or median is adaptively selected as the robustness center estimate. According to the instantaneous electrical interference intensity and temperature stability reflected by the auxiliary diagnostic parameters, the acceptable fluctuation threshold used to evaluate data quality is dynamically adjusted, and a composite verification is performed on the robustness center estimate and the statistical distribution of the second dataset based on this threshold. Only when the composite verification passes is a zero-point reference write operation including post-write verification executed, and a calibration traceability record containing all key parameters and processes is generated. If the verification fails, the calibration is interrupted and a fault warning containing specific diagnostic information is output.

[0082] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for online maintenance of pressure transmitters, characterized in that, Includes the following steps: S1: Perform zero-point calibration on the pressure transmitter to be inspected, and simultaneously start a sampling window of a predetermined duration and a background noise monitoring window; S2: Within the sampling window, continuously acquire the raw pressure signal value of the pressure transmitter at the first sampling frequency to form the first dataset; Within the background noise monitoring window, monitor and record auxiliary diagnostic parameters related to signal integrity within the pressure transmitter, specifically including: Within the background noise monitoring window, the voltage fluctuation signal at the reference voltage terminal of the analog-to-digital converter of the pressure transmitter and the reading change signal of the internal temperature sensor of the pressure transmitter within the monitoring window are simultaneously acquired. The effective value of the voltage fluctuation signal and the average rate of change of the reading change signal are calculated to form auxiliary diagnostic parameters with two dimensions. S3: Perform a physical validity test on the first dataset based on the continuity of the pressure signal, remove outlier data segments that do not conform to the physical change pattern, generate the second dataset, and calculate the robust center estimate of the second dataset, specifically including: All pressure data points in the second dataset are sequentially numbered based on time, and all pressure data points are treated as a time-domain signal sequence for frequency domain transformation analysis to obtain the spectral distribution of the time-domain signal sequence; Analyze the main frequency components of energy concentration in the spectrum distribution and identify whether they fall within the pre-defined reasonable frequency band of the physical process; If the main frequency components fall within a reasonable frequency band, the second dataset is determined to reflect the real physical process, and the arithmetic mean of the second dataset is used as the robust center estimate; if the main frequency components fall outside a reasonable frequency band, the median of the second dataset is used as the robust center estimate. S4: Based on the status of the auxiliary diagnostic parameters, adjust the acceptable fluctuation threshold used to assess the quality of the second dataset, and based on the adjusted acceptable fluctuation threshold, perform a composite verification of the robust center estimate and the statistical distribution of the second dataset, specifically including: The interference assessment value, which characterizes the instantaneous electrical interference intensity, is extracted from the auxiliary diagnostic parameters, and the interference assessment value is used to amplify the preset benchmark threshold to obtain the first correction threshold. The rate of change characterizing temperature stability is extracted from the auxiliary diagnostic parameters, and the rate of change is used to scale the first correction threshold to generate the final acceptable fluctuation threshold. Simultaneously verify whether the robust center estimate is within the range centered on the average of the second dataset and with the acceptable fluctuation threshold as the radius, and whether the range of the second dataset is less than a preset multiple of the acceptable fluctuation threshold. When both conditions are met, the composite verification is passed. S5: Only when the composite verification passes, the robust center estimate is written to the zero-point storage unit of the pressure transmitter to complete the calibration, and a calibration traceability record containing auxiliary diagnostic parameters, acceptable fluctuation thresholds and verification results is generated; if the verification fails, the calibration process is interrupted and a fault warning containing auxiliary diagnostic parameters is output.

2. The method for online maintenance of a pressure transmitter according to claim 1, characterized in that, The formation of the first dataset specifically includes: The sampling window is divided into multiple consecutive and partially overlapping data frames; Within each data frame, the acquired raw pressure signal value is processed in real time using a sliding median filter to obtain the corresponding representative value within the frame; Collect the intra-frame representative value sequence of all data frames and calculate the mean absolute difference of the intra-frame representative value sequence; If the mean absolute difference is less than the preset first threshold, all intra-frame representative values ​​are stored as the first dataset; if the mean absolute difference is greater than or equal to the first threshold, the current data is discarded and the sampling window is automatically extended for a predetermined duration. The acquisition and division steps are then repeated until the obtained intra-frame representative value sequence satisfies the condition that the mean absolute difference is less than the first threshold.

3. The method for online maintenance of pressure transmitters according to claim 1, characterized in that, S3 specifically includes: Calculate the rate of pressure change between adjacent data points in the first dataset to form a rate of change sequence; Identify abrupt changes in the rate of change sequence where the absolute value exceeds a preset physical threshold, and mark each abrupt change and a predetermined number of data points before and after it as the segment to be inspected; Check whether the pressure value of each data point in the section to be inspected exhibits an oscillation or step pattern that does not conform to the law of inertia; All data from the inspected sections that were determined to conform to the laws of physical change were re-integrated to form a continuous and physically reliable second dataset.

4. The method for online maintenance of a pressure transmitter according to claim 3, characterized in that, The process of checking whether the pressure value of each data point in the inspected section exhibits an oscillation or step pattern that does not conform to the law of inertia specifically includes: Calculate the absolute value sequence of the first difference of pressure data points within the section to be inspected; Identify all local maxima in the absolute value sequence, calculate the average amplitude of these local maxima, and determine whether there are three or more consecutive local maxima whose amplitudes all exceed a set multiple of the average amplitude. If it exists, the corresponding segment is determined to exhibit an oscillation pattern that does not conform to the law of inertia; If it does not exist, further check whether the pressure difference between the start and end points of the corresponding section is greater than another set multiple of the average amplitude and whether the direction of change is contrary to the direction of the physical process, in order to determine whether it is a step form that does not conform to the law.

5. The method for online maintenance of a pressure transmitter according to claim 1, characterized in that, The process of obtaining the first correction threshold specifically includes: Obtain the voltage ripple signal from the auxiliary diagnostic parameters, and calculate the root mean square value of the voltage ripple signal within a predetermined short time window as the original interference intensity; The zero-crossing interval sequence of the voltage ripple signal is analyzed, and the coefficient of variation of the zero-crossing interval sequence interval value is statistically analyzed to characterize the suddenness and irregularity of the interference. The original interference intensity and the coefficient of variation are weighted and fused to generate the interference assessment value; The first correction threshold is obtained by multiplying the interference evaluation value by a preset baseline threshold.

6. An online maintenance system based on pressure transmitters, characterized in that, A method for performing online maintenance of a pressure transmitter as described in any one of claims 1-5 includes: Synchronous trigger module: performs zero-point calibration on the pressure transmitter to be inspected, and simultaneously starts a sampling window of a predetermined duration and a background noise monitoring window; Dual-channel parallel acquisition module: Within the sampling window, the original pressure signal value of the pressure transmitter is continuously acquired at the first sampling frequency to form the first dataset; within the background noise monitoring window, auxiliary diagnostic parameters related to signal integrity inside the pressure transmitter are monitored and recorded. Physical validity test and robust estimation module: Performs a physical validity test on the first dataset based on the continuity of the pressure signal, removes abnormal data segments that do not conform to the physical change law, generates the second dataset, and calculates the robust center estimate of the second dataset; Dynamic threshold composite verification module: Based on the status of auxiliary diagnostic parameters, adjust the acceptable fluctuation threshold used to evaluate the quality of the second dataset, and perform composite verification on the robust center estimate and the statistical distribution of the second dataset based on the adjusted acceptable fluctuation threshold; Calibration execution and traceability record module: Only when the composite verification passes, the robust center estimate is written to the zero-point storage unit of the pressure transmitter to complete the calibration, and a calibration traceability record containing auxiliary diagnostic parameters, acceptable fluctuation thresholds and verification results is generated; if the verification fails, the calibration process is interrupted and a fault warning containing auxiliary diagnostic parameters is output.

Images