A mechanical error calibration method and system for an electronic water meter, a terminal and a storage medium

Abstract

This invention discloses a method, system, terminal, and storage medium for calibrating the mechanical error of an electronic water meter. The method includes: determining multiple key parameters of water flow in the electronic water meter and collecting water flow data to obtain multiple target water flow data; performing data preprocessing and feature extraction on the multiple target water flow data to obtain multiple target water flow features; calculating the predicted water flow corresponding to each target water flow feature, and calculating multiple water flow error coefficients based on the multiple predicted water flow and actual water flow; obtaining the parameter weights corresponding to each key water flow parameter, and correcting each water flow error coefficient to obtain multiple correction parameters; adjusting the key water flow parameters according to each correction parameter to obtain the adjustment result and the target water flow result. This invention, by calculating the water flow error coefficients and then adjusting the parameters of the electronic water meter based on the water flow error coefficients, can effectively improve the accuracy of the water flow results.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method, system, terminal, and computer-readable storage medium for calibrating the mechanical errors of an electronic water meter. Background Technology

[0002] Currently, electronic water meters generally use ultrasonic water meters, which are water meters that use ultrasonic technology to measure flow rate and have advantages such as high accuracy and good stability. With the rapid development and popularization of Internet of Things (IoT) technology, ultrasonic water meters, as an important component of smart city construction, have been widely used in urban water supply systems.

[0003] However, ultrasonic water meters are easily affected by the environment and their own mechanical structure during practical applications, which can lead to measurement errors and inaccurate water flow measurement results.

[0004] Therefore, existing technologies still need to be improved and developed. Summary of the Invention

[0005] The main objective of this invention is to provide a mechanical error calibration method, system, terminal, and computer-readable storage medium for electronic water meters. This invention aims to solve the problem that ultrasonic water meters are easily affected by the environment and their own mechanical structure during practical applications, resulting in measurement errors and inaccurate water flow measurement results.

[0006] To achieve the above objectives, the present invention provides a method for calibrating the mechanical error of an electronic water meter, the method comprising the following steps: Multiple key parameters of water flow in the electronic water meter are determined, and water flow data is collected based on each of the key parameters to obtain multiple target water flow data. Data preprocessing and feature extraction are performed on multiple target water flow data to obtain multiple target water flow features; Calculate the predicted flow rate corresponding to each of the target water flow characteristics, and calculate multiple water flow error coefficients based on the multiple predicted flow rates and the actual flow rates; Obtain the parameter weight corresponding to each of the key parameters of water flow, and perform correction processing on each of the water flow error coefficients according to the parameter weights to obtain multiple correction parameters; The corresponding key parameters of water flow in the electronic water meter are adjusted according to each of the correction parameters to obtain the adjustment results and the target water flow results.

[0007] Optionally, the mechanical error calibration method for the electronic water meter, wherein determining multiple key parameters of water flow of the electronic water meter, and collecting water flow data based on each of the key parameters of water flow to obtain multiple target water flow data, specifically includes: Several key parameters affecting the water flow of electronic water meters were determined, including temperature parameters, water meter mechanical parameters, and fluid condition parameters. Water flow data is collected within a preset time period based on the temperature parameter, the water meter mechanical parameter, and the fluid condition parameter, respectively, to obtain multiple target water flow data.

[0008] Optionally, the mechanical error calibration method for the electronic water meter, wherein the step of performing data preprocessing and feature extraction processing on multiple target water flow data to obtain multiple target water flow features specifically includes: Data filtering, integrity detection, and missing value filling are performed on multiple target water flow data to obtain multiple preprocessed data. A preset deep learning model is used to extract features from multiple preprocessed data to obtain multiple target water flow features.

[0009] Optionally, in the mechanical error calibration method for the electronic water meter, the target water flow characteristics include temperature characteristics, water meter mechanical characteristics, and fluid operating condition characteristics; the water volume error coefficient includes a first water volume error coefficient, a second water volume error coefficient, and a third water volume error coefficient. The calculation of the predicted flow rate corresponding to each of the target water flow characteristics, and the calculation of multiple water flow error coefficients based on the multiple predicted flow rates and the actual flow rates, specifically includes: Determine the actual water flow of the electronic water meter; Calculate the first predicted water flow corresponding to the temperature characteristic, and calculate the difference between the first predicted water flow and the actual water flow to obtain the first water flow error coefficient; Calculate the second predicted water flow corresponding to the mechanical characteristics of the water meter, and calculate the difference between the second predicted water flow and the actual water flow to obtain the second water flow error coefficient; Calculate the third predicted water flow rate corresponding to the fluid operating condition characteristics, and calculate the difference between the third predicted water flow rate and the actual water flow rate to obtain the third water flow error coefficient.

[0010] Optionally, the mechanical error calibration method for the electronic water meter, wherein obtaining the parameter weights corresponding to each of the key water flow parameters, and correcting each water flow error coefficient according to the parameter weights to obtain multiple correction parameters, specifically includes: Obtain historical water flow data from the electronic water meter, and obtain the parameter weight corresponding to each of the key parameters of water flow based on the historical water flow data; Based on the parameter weights, the first water volume error coefficient, the second water volume error coefficient, and the third water volume error coefficient are corrected to obtain the correction parameters corresponding to each key water flow parameter.

[0011] Optionally, the mechanical error calibration method for the electronic water meter, wherein adjusting the corresponding key water flow parameter in the electronic water meter according to each correction parameter to obtain the adjustment result and the target water flow result specifically includes: The temperature parameter, mechanical parameter, and fluid condition parameter in the electronic water meter are adjusted according to the correction parameters until the temperature parameter, mechanical parameter, and fluid condition parameter are all within the corresponding preset parameter range, thus obtaining the adjustment result. Based on the adjustment results, the water flow data of the electronic water meter is collected to obtain the target water flow result.

[0012] Optionally, the mechanical error calibration method for the electronic water meter includes data filtering processing including threshold filtering, sliding filtering, and Kalman filtering.

[0013] Furthermore, to achieve the above objectives, the present invention also provides a mechanical error calibration system for an electronic water meter, wherein the mechanical error calibration system for the electronic water meter includes: The water flow data acquisition module is used to determine multiple key parameters of water flow of the electronic water meter, and to acquire water flow data based on each of the key parameters of water flow to obtain multiple target water flow data. The water flow data processing module is used to perform data preprocessing and feature extraction on multiple target water flow data to obtain multiple target water flow features; The water volume error coefficient calculation module is used to calculate the predicted water flow corresponding to each of the target water flow characteristics, and to calculate multiple water volume error coefficients based on the multiple predicted water flow and the actual water flow. The water volume error coefficient correction module is used to obtain the parameter weights corresponding to each of the key parameters of water flow, and to correct each of the water volume error coefficients according to the parameter weights to obtain multiple correction parameters. The water flow result output module is used to adjust the corresponding key parameters of water flow in the electronic water meter according to each of the correction parameters, so as to obtain the adjustment result and the target water flow result.

[0014] Furthermore, to achieve the above objectives, the present invention also provides a terminal, wherein the terminal includes: a memory, a processor, and a mechanical error calibration program for an electronic water meter stored in the memory and executable on the processor, wherein when the mechanical error calibration program for the electronic water meter is executed by the processor, it implements the steps of the mechanical error calibration method for the electronic water meter as described above.

[0015] Furthermore, to achieve the above objectives, the present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a mechanical error calibration program for an electronic water meter, and when the mechanical error calibration program for the electronic water meter is executed by a processor, it implements the steps of the mechanical error calibration method for the electronic water meter as described above.

[0016] In this invention, multiple key parameters for water flow in an electronic water meter are determined, and water flow data is collected based on each key parameter to obtain multiple target water flow data. Data preprocessing and feature extraction are performed on the multiple target water flow data to obtain multiple target water flow features. The predicted water flow corresponding to each target water flow feature is calculated, and multiple water flow error coefficients are calculated based on the predicted and actual water flow. The parameter weights corresponding to each key water flow parameter are obtained, and each water flow error coefficient is corrected based on the parameter weights to obtain multiple correction parameters. The corresponding key water flow parameters in the electronic water meter are adjusted based on each correction parameter to obtain the adjustment result and the target water flow result. This invention effectively improves the accuracy of water flow results by calculating water flow error coefficients based on predicted and actual water flow, and then adjusting the parameters of the electronic water meter based on these error coefficients. Attached Figure Description

[0017] Figure 1 This is a flowchart of a preferred embodiment of the mechanical error calibration method for the electronic water meter of the present invention; Figure 2 This is a structural diagram of a preferred embodiment of the mechanical error calibration system for the electronic water meter of the present invention; Figure 3 This is a structural diagram of a preferred embodiment of the terminal of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] Currently, electronic water meters generally use ultrasonic water meters, which are water meters that use ultrasonic technology to measure flow rate and have advantages such as high accuracy and good stability. With the rapid development and popularization of Internet of Things (IoT) technology, ultrasonic water meters, as an important component of smart city construction, have been widely used in urban water supply systems.

[0020] However, ultrasonic water meters are easily affected by the environment and their own mechanical structure during practical applications, which can lead to measurement errors and inaccurate water flow measurement results.

[0021] To address the aforementioned problems, this invention proposes a mechanical error calibration method for electronic water meters. By calculating the water flow error coefficient based on the predicted and actual water flow, the parameters of the electronic water meter are adjusted according to the water flow error coefficient, which can effectively improve the accuracy of the water flow results.

[0022] The preferred embodiment of the present invention describes a mechanical error calibration method for electronic water meters, such as... Figure 1 As shown, the mechanical error calibration method for the electronic water meter includes the following steps: Step S10: Determine multiple key parameters of water flow of the electronic water meter, and collect water flow data according to each of the key parameters to obtain multiple target water flow data.

[0023] During the use of electronic water meters, they are easily affected by the external environment and internal mechanical mechanisms, which may cause deviations in measurement accuracy, resulting in inaccurate water flow readings. Therefore, it is necessary to monitor electronic water meters in real time and adjust key parameters affecting water flow, such as temperature, mechanical parameters, and fluid condition parameters, to improve the measurement accuracy of electronic water meters. In this invention, the target water flow data is water flow data collected within a preset time period based on each key parameter.

[0024] This invention combines the structure of the electronic water meter's metering mechanism, the operating conditions of the fluid pipeline, and the interference characteristics of the on-site environment. It divides the key parameters affecting the water flow of the electronic water meter into three core dimensions: temperature parameters, water meter mechanical parameters, and fluid condition parameters. Among them, the temperature parameters cover three sub-indicators: water medium temperature, water meter main control circuit board operating temperature, and metering chamber ambient temperature. The water meter mechanical parameters include five major mechanical operation indicators: metering impeller rotation pulse frequency, mechanism transmission clearance, check valve closing response time, mechanism wear and deformation degree, and water meter installation tilt angle. The fluid condition parameters include six major fluid operation indicators: pipeline instantaneous water pressure, water pressure fluctuation amplitude, pipeline air content, instantaneous water flow velocity, water flow direction, and pipeline turbulence.

[0025] To address the three types of differentiated key parameters mentioned above, this invention establishes an integrated multi-source synchronous acquisition hardware architecture for electronic water meters. Relying on the water meter's built-in MCU main control unit, high-precision sensor array, and signal filtering and conditioning module, it performs time-aligned and high-frequency synchronized full-domain water flow data acquisition within a preset complete water use cycle (covering four typical operating conditions: peak water use, low water use, water outage and resumption, and pipeline idling) based on the temperature parameters, the water meter's mechanical parameters, and the fluid operating parameters.

[0026] Specifically, several key parameters affecting the water flow of the electronic water meter are determined, including temperature parameters, water meter mechanical parameters, and fluid condition parameters. Water flow data are collected within a preset time period based on the temperature parameters, water meter mechanical parameters, and fluid condition parameters to obtain multiple target water flow data.

[0027] The process of collecting water flow data for temperature parameters is as follows: This invention adopts a combination of surface-mount NTC thermistors and invasive temperature sensors, which are respectively attached to the electronic water meter circuit board, the inside of the metering cavity, and the inner wall of the inlet pipe. At a high frequency of 100ms / time, the circuit operating temperature, medium water temperature, and cavity ambient temperature time sequence data are collected synchronously. Furthermore, the synchronous linkage electronic water meter's built-in flow metering chip collects instantaneous flow pulse data and cumulative water flow raw data at the corresponding temperature node in real time, eliminates electrical signal noise interference caused by temperature drift, and completes the time-series binding of temperature parameters and water flow metering data.

[0028] This invention can accurately capture the temperature drift error of water meter electronic components and the measurement deviation caused by the change of water viscosity with temperature under high and low temperature environments. It solves the industry pain points of water meter under-measurement caused by the increase of water viscosity at low temperature and water meter over-measurement caused by the deformation of the plastic structure of the mechanism at high temperature. It also eliminates the problem of distortion of original water flow data caused by temperature nonlinear interference, and ensures that the quantification accuracy of temperature interference factor is improved by more than 20% when modeling subsequent errors.

[0029] The water flow data acquisition process for the mechanical parameters of the water meter is as follows: This invention relies on a built-in Hall pulse sensor, tilt detection module, and mechanical displacement sensing unit to collect water meter mechanical operating status data in a completely contactless manner. It continuously collects the pulse output signal of the metering impeller to count the actual rotation frequency of the impeller and compares it with the standard theoretical pulse value to obtain the transmission clearance deviation of the mechanism. It monitors the installation tilt angle of the water meter housing in real time and records the offset from the standard horizontal installation position. It detects the opening and closing displacement data of the check valve and counts the number of backflows due to check valve failure under water pressure fluctuations. It synchronously collects water metering data for the corresponding mechanical state every second, distinguishes between two types of water flow data: normal water flow measurement and mechanical idling virtual measurement, and realizes a one-to-one correspondence between mechanical faults and abnormal water flow.

[0030] This invention enables full data traceability of hidden mechanical defects in water meters (such as wear of the movement, installation tilt, and failure of the check valve), accurately separates invalid idle water volume caused by abnormal mechanical structure from actual water usage, and makes up for the shortcomings of traditional electronic water meters that only collect flow data and cannot distinguish abnormal water flow due to mechanical failure. It provides complete mechanical dimension raw data support for subsequent elimination of mechanical errors and correction of water meter measurement benchmark values.

[0031] The process of collecting water flow data for fluid operating parameters is as follows: In this invention, high-precision pressure sensors, gas detection sensors, and flow velocity acquisition modules are respectively deployed in the straight pipe section before the water meter and at the water outlet after the meter. They are synchronized with the water meter's native flow acquisition channel to synchronously collect data on pipeline static pressure, dynamic water pressure fluctuations, air bubble content in the pipe, instantaneous flow velocity and direction, and pipeline turbulence disturbances. This fully covers all operating scenarios, including morning and evening network water pressure fluctuations, air accumulation in the water supply pipeline after water outages, bidirectional backflow of water in the pipeline, and turbulent flow due to valve throttling. Multivariable fluid time-series data are continuously collected within a preset time period, and the actual water flow rate measured by the water meter is synchronously bound to each set of fluid operating conditions.

[0032] This invention fully restores the real operating state of water meters under complex fluid conditions in pipeline networks. It can accurately capture incremental data of water meter idling caused by air accumulation in pipelines, water pressure fluctuations, and water backflow, solving the problem that single flow data cannot explain abnormal water flow under operating conditions. At the same time, it realizes the coupling correlation between multiple fluid parameters and water flow, providing high-fidelity, time-aligned target water flow data for subsequent analysis of the influence weight of fluid conditions on the metering error of electronic water meters.

[0033] In summary, this invention ultimately integrates multiple sets of raw water flow data obtained from the acquisition of three types of parameters, including time-series matching, noise filtering, and operating condition labeling, to obtain multiple target water flow data with complete dimensions, unified time series, and traceable error sources. This provides reliable and comprehensive underlying data support for subsequent analysis of water flow error in electronic water meters under multi-parameter coupling, feature extraction, and the construction of metering error compensation models.

[0034] Step S20: Perform data preprocessing and feature extraction on the multiple target water flow data to obtain multiple target water flow features.

[0035] After obtaining the target flow data, it is necessary to preprocess the data because it may contain abnormal or irrelevant data. This preprocessing includes data filtering, integrity checks, and missing value imputation. Further, feature extraction is required from the preprocessed data to obtain the key features affecting the target flow data (i.e., target flow features). These key features can then be used to calculate the predicted flow rate.

[0036] Understandably, for multiple time-aligned, multi-source coupled target water flow data collected, this invention addresses inherent data defects in the electronic water meter's on-site operating environment, such as electromagnetic interference, instantaneous pulse noise in the pipeline network, short-term signal interruptions in sensors, and abrupt changes in operating conditions. It sequentially performs data filtering and noise reduction, data integrity detection, and intelligent missing value filling on multiple target water flow data sets. This process removes invalid noise, outliers, and incomplete data from the original water flow data, unifies the time-series format and accuracy of multi-source parameter data, and ultimately yields multiple pre-processed data sets with standardized format, extremely low noise, and complete time series. Specifically, to address the fluctuation characteristics of multi-source heterogeneous water flow data from water meters, the data filtering process employs a three-stage progressive filtering scheme, including threshold filtering, moving average filtering, and Kalman filtering combined for noise reduction. These three types of filtering are tailored to different noise scenarios encountered by water meters.

[0037] Specifically, multiple target water flow data are subjected to data filtering, integrity detection, and missing value imputation to obtain multiple preprocessed data; the data filtering includes threshold filtering, sliding filtering, and Kalman filtering; a preset deep learning model is used to extract features from the multiple preprocessed data to obtain multiple target water flow features.

[0038] First, this invention combines the national standard flow range of electronic water meters with the normal fluctuation range of on-site pipeline water pressure and medium temperature, and pre-sets upper and lower limit thresholds for various parameters such as flow rate, pressure, temperature, and impeller pulse. It then traverses all target water flow data and directly removes abnormal data points that exceed the physical operating limits (such as instantaneous large flow wild values ​​caused by instantaneous water hammer impact in the pipeline or abnormal pulse jump values ​​caused by electromagnetic interference in the circuit), retaining valid water flow data within the normal physical operating range of the water meter.

[0039] This invention effectively avoids interference from extreme outliers in subsequent filtering and feature extraction models by quickly stripping away physically meaningless extreme abnormal data caused by sudden on-site interference. It reduces the impact of invalid abnormal data on the results of metering error analysis from the source and is suitable for sudden on-site interference scenarios such as water hammer impact on water meters and electromagnetic interference from power grids.

[0040] Furthermore, this invention employs a fixed-window-length moving average filtering algorithm. For the continuous time-series water flow data remaining after threshold filtering, the algorithm uses 10 adjacent sets of time-series data as a sliding window to calculate the average values ​​of multiple parameters such as flow rate, water pressure, and temperature within each window, replacing the original data in the center of the window. This is used to smooth out high-frequency random small noise generated during sensor acquisition and eliminate the problem of slight jitter in water flow data caused by micro-bubbles in the pipe and background noise of the sensor circuit.

[0041] The high-frequency spikes in the smooth time-series water flow curve can restore the true and continuous change trend of water flow in the pipe network, and solve the problem of flow time-series curve jitter caused by the inherent background noise of the electronic water meter sensing unit. This invention ensures the continuity of time-series data, while preserving the true slow fluctuation characteristics of water flow, without destroying the effective operating condition data.

[0042] Furthermore, this invention relies on Kalman filtering to construct the state-space equation of water meter flow data, and combines the physical laws of water flow operation to establish observation equations and prediction equations. It performs optimal estimation of the time series data after the first two stages of filtering, dynamically compensates for the time series deviation caused by sensor sampling delay and asynchronous acquisition of multi-source parameters, corrects the dynamic error of data under dynamic conditions of water pressure fluctuation and water flow backflow, and realizes the time series synchronous calibration of multi-source water flow data.

[0043] This invention overcomes the shortcomings of the previous two static filters in adapting to dynamic working conditions, thereby accurately compensating for the timing misalignment error caused by asynchronous acquisition of multiple sensors. It can effectively adapt to dynamic and complex working conditions such as dynamic fluctuations in pipeline water pressure, water outage and resumption, and bidirectional water flow backflow, further improving the timing consistency and data authenticity of multi-source water flow data, and improving the overall data noise reduction accuracy by more than 35%.

[0044] After completing the three-stage filtering, this invention performs integrity verification on a time-series basis for three types of synchronously collected water flow data: temperature, mechanical parameters, and fluid conditions. On one hand, it detects whether there are single-parameter missing or partially garbled issues in the multi-dimensional parameters at a single time-series node; on the other hand, it detects whether there are fragmented data packet loss or sampling interruptions in the entire time-series. Simultaneously, it marks the location, duration, and type of missing parameters, distinguishing between short-term, single-point random missing data and long-term, continuous missing data segments, and generates a data integrity detection report.

[0045] This invention can accurately locate various data incompleteness problems caused by short-term power failure of water surface sensors, signal transmission interruption, and wireless communication packet loss, and realize the fine classification of missing data, providing a basis for subsequent differential missing value filling strategies, and avoiding the problems of feature extraction failure and model overfitting caused by directly inputting incomplete data into deep learning models.

[0046] Furthermore, based on the integrity detection results, this invention adopts a differentiated intelligent filling strategy: for short-term, single-point, randomly missing data, it uses interpolation of adjacent time-series data to complete rapid filling; for long-term, continuously missing data segments, it relies on similar water flow data from the water meter's historical operating conditions and combines it with current real-time operating parameters to construct a time-series fitting model for adaptive filling; after filling, it re-verifies the data fluctuation amplitude to ensure that the filled data conforms to the actual water flow change pattern of the water meter and that there is no artificially introduced false data deviation.

[0047] This invention can completely supplement incomplete water flow data caused by communication anomalies and short-term sensor failures, ensuring that the entire time-series water flow data is uninterrupted and has a uniform time-series length. It eliminates the need to directly delete incomplete and valid time-series segments, maximizing the preservation of water use data throughout the entire cycle, thereby ensuring that the input data of subsequent deep learning models has a uniform dimension and complete time series.

[0048] Furthermore, this invention inputs standardized preprocessed data (after filtering, integrity detection, and missing value imputation) into a pre-defined hybrid deep learning model (one-dimensional CNN + bidirectional GRU hybrid feature extraction model) adapted to multivariate time-series data of water meters. This model performs fully automated extraction of shallow local features, mid-level time-dependent features, and high-level multi-parameter coupled features, comprehensively uncovering the implicit correlation between three types of key parameters and abnormal water flow in electronic water meters. The specific extraction process is as follows: 1. Shallow local spatial feature extraction (one-dimensional CNN module): By using one-dimensional convolutional layers and pooling layers, local sliding convolution operations are performed on the preprocessed multivariate time-series water flow data to automatically extract local fine-grained features within a short time window, including short-term water pressure pulse fluctuation features, water temperature change features, instantaneous abnormality features of the core pulse, and short-term impact features of gas in the pipeline, as well as other local operating condition change features.

[0049] 2. Mid-level long-term time-dependent feature extraction (bidirectional GRU module): The local features output by the CNN are input into the bidirectional gated recurrent unit to mine the long-term dependencies of water flow data from both forward and reverse time series. This captures long-term slow abnormal water flow features such as the periodic changes in water usage peaks and troughs throughout the day, the gradual changes in slow gas accumulation in pipes, and the slow deterioration of the mechanism over a long period of time. This makes up for the deficiency of a single convolutional network in capturing long-term temporal correlations.

[0050] 3. High-level multi-parameter coupled feature fusion: The fully connected layer is used to fuse and reduce the dimensionality of three heterogeneous features: temperature parameters, water meter mechanical parameters, and fluid condition parameters. Redundant and duplicate features are eliminated, and core sensitive features that are highly correlated with abnormal water flow are selected. Finally, multiple target water flow features with unified dimensions and high recognizability are output.

[0051] In summary, this invention overcomes the drawbacks of traditional manual feature extraction, which relies on expert experience and suffers from one-sided features. It automatically mines explicit and implicit metering error features under multi-parameter coupling. It can accurately distinguish abnormal water flow characteristics caused by four different reasons: pipeline air accumulation and idling, water pressure fluctuation and idling, mechanical failure of the mechanism, and temperature drift. This effectively improves the accuracy of subsequent water meter metering error tracing, abnormal water flow identification, and metering compensation model. The overall abnormal water flow identification accuracy is improved by more than 28%, making it suitable for abnormal water flow detection needs in all scenarios of electronic water meters.

[0052] Step S30: Calculate the predicted water flow rate corresponding to each of the target water flow characteristics, and calculate multiple water flow error coefficients based on the multiple predicted water flow rates and the actual water flow rates. The target water flow characteristics include temperature characteristics, water meter mechanical characteristics, and fluid operating condition characteristics; the water flow error coefficients include a first water flow error coefficient, a second water flow error coefficient, and a third water flow error coefficient.

[0053] This invention accurately determines the actual water flow rate of the electronic water meter within a preset time period, under conditions of no external interference and no correction for operating conditions, based on the obtained target water flow characteristics in terms of temperature, mechanical properties, and fluid conditions, combined with the actual metering readings returned by the built-in metering reference module of the electronic water meter. This actual water flow rate is the original cumulative water consumption directly collected by the water meter's metering mechanism without error compensation, serving as the true reference value for subsequent error comparison. Subsequently, a single-factor independent regression prediction algorithm is used to predict the water flow rate in different dimensions based on the three types of differentiated water flow characteristics. The metering deviation caused by each interference factor is quantified and decomposed one by one, and the water volume error coefficient of the corresponding dimension is calculated.

[0054] Specifically, the actual water flow of the electronic water meter is determined; the first predicted water flow corresponding to the temperature characteristic is calculated, and the difference between the first predicted water flow and the actual water flow is calculated to obtain a first water flow error coefficient; the second predicted water flow corresponding to the mechanical characteristic of the water meter is calculated, and the difference between the second predicted water flow and the actual water flow is calculated to obtain a second water flow error coefficient; the third predicted water flow corresponding to the fluid operating condition characteristic is calculated, and the difference between the third predicted water flow and the actual water flow is calculated to obtain a third water flow error coefficient.

[0055] First, this invention retrieves the temperature target water flow features extracted by the aforementioned deep learning model. These features include three core representations: water temperature temporal variation features, circuit board temperature drift features, and metering cavity thermal deformation features. The temperature features are then input into a pre-trained temperature measurement error regression prediction model. The model relies on the mapping relationship between water temperature and water viscosity, the temperature drift law of components, and the thermal expansion and contraction deformation law of plastic mechanisms to fit and calculate the first predicted water flow rate theoretically output by the electronic water meter when only temperature is affected. This predicted value is the simulated water flow rate of the water meter under the influence of pure temperature factors after eliminating interference from mechanical structure and fluid conditions.

[0056] Furthermore, a difference calculation is performed, subtracting the actual water flow from the electronic water meter from the first predicted water flow, and normalizing the difference result into a dimensionless form to eliminate the data magnitude differences caused by different water flow ranges and different sampling periods. Finally, the first water flow error coefficient, which characterizes the degree of temperature interference, is obtained.

[0057] This invention achieves independent quantitative separation of the impact of temperature as a single interference factor on water meter flow rate, avoiding the coupling and confusion between temperature error, mechanical error, and fluid condition error. It accurately quantifies the degree of metering deviation caused by high-temperature core deformation, low-temperature water viscosity changes, and circuit temperature drift, and intuitively reflects the amplitude of water meter temperature measurement deviation across the entire temperature range, providing accurate quantitative basis for subsequent separate compensation based on temperature.

[0058] Furthermore, this invention retrieves the extracted mechanical target water flow characteristics of the water meter, encompassing four major mechanical anomaly characteristics: impeller rotation pulse deviation, core transmission clearance wear, check valve backflow idling, and water meter installation tilt angle offset. Then, the water meter's mechanical characteristics are input into a dedicated mechanical metering error prediction model. This model isolates external environmental variables such as temperature and fluid conditions, relying solely on the water meter's own mechanical structure and operating state to calculate a second predicted water flow rate affected only by mechanical aging, installation deviation, and valve failure. The difference between the second predicted water flow rate and the actual water flow rate of the electronic water meter is calculated, and error normalization calibration is performed based on the water meter's rated flow rate specification. This eliminates differences in mechanical reference values ​​between water meters of different diameters, ultimately yielding a second water flow error coefficient characterizing the degree of interference from mechanical structural defects.

[0059] This invention, by independently dissecting internal mechanical faults such as wear and tear of the water meter's movement, check valve failure, and installation tilt, can accurately distinguish the inherent measurement deviation caused by the aging of the water meter's own hardware. It solves the problem that traditional water meters cannot distinguish between internal mechanical errors and external operating condition errors. It can directly determine the degree of aging of the water meter's movement and the level of mechanical faults based on the second water volume error coefficient, supporting the water meter's fault self-check and early warning function.

[0060] Furthermore, this invention retrieves the completed target water flow characteristics of the fluid operating conditions. These characteristics include features such as network water pressure fluctuations, air bubble disturbances in the pipeline, bidirectional backflow of water, and turbulent flow interference in the pipeline, representing network fluid disturbances. Then, the fluid operating condition characteristics are input into the fluid disturbance metering prediction model. This model shields the inherent deviations of the temperature environment and the water meter's mechanical structure, only restoring external fluid operating condition disturbances such as network water pressure fluctuations, air accumulation during water outages and resuming supply, water hammer impacts, and backflow of water. The model then calculates the third predicted water flow of the electronic water meter under pure fluid operating condition disturbances. Finally, the absolute deviation between the third predicted water flow and the actual water flow of the electronic water meter is calculated. Error correction is then performed in segments based on time-series water usage conditions, avoiding differences in error base under different flow rates during peak and off-peak water usage. After normalization, a third water volume error coefficient characterizing the degree of influence of network fluid disturbances is obtained.

[0061] This invention can accurately quantify the problems of water meter idling and over-metering caused by complex fluid conditions in the pipeline network, isolate the metering errors caused by the external pipeline environment, and clearly distinguish between the actual water consumption and the false water flow caused by pipeline disturbances. At the same time, it can judge the severity of the on-site pipeline conditions based on the magnitude of the third water volume error coefficient, and is suitable for complex on-site application scenarios such as high-rise secondary water supply and large water pressure fluctuations in old pipeline networks.

[0062] In summary, this invention achieves error decoupling of three types of interference factors through the aforementioned multidimensional independent prediction and error solving methods, obtaining one-to-one corresponding first water volume error coefficient, second water volume error coefficient, and third water volume error coefficient. The three types of error coefficients are independent of each other and have no coupling interference, which can completely restore the component composition of the overall measurement error of the electronic water meter, providing complete quantitative parameter support for subsequent multidimensional integrated measurement error compensation and accurate correction of the final measurement reading of the water meter.

[0063] Step S40: Obtain the parameter weight corresponding to each of the key parameters of water flow, and perform correction processing on each of the water flow error coefficients according to the parameter weights to obtain multiple correction parameters.

[0064] To eliminate the differentiated impact of temperature parameters, water meter mechanical parameters, and fluid operating condition parameters on the water flow error of electronic water meters under different application scenarios, and to solve the problems of single-dimensional water flow error coefficients not considering the proportion of operating condition influence and one-sided error assessment, the present invention sets up the following steps: First, acquire massive historical water flow data stored locally in the electronic water meter and synchronized in the cloud of the water affairs platform. Then, based on the historical time series data, complete the quantitative solution of the influence weight of the three types of key parameters. Subsequently, based on the precise weight, the three types of original water flow error coefficients are weighted and corrected to weaken the error interference of secondary interference factors and amplify the influence of core error factors, and finally obtain various correction parameters that fit the actual on-site operating conditions.

[0065] Specifically, historical water flow data of the electronic water meter is obtained, and parameter weights corresponding to each of the key water flow parameters are obtained based on the historical water flow data; the first water flow error coefficient, the second water flow error coefficient, and the third water flow error coefficient are corrected based on the parameter weights to obtain correction parameters corresponding to each of the key water flow parameters.

[0066] The detailed process for obtaining the weights of key parameters related to water flow based on historical water flow data in this invention is as follows: 1. First, the historical water flow data stored in the cloud and the local FLASH storage module of the water meter is targeted and filtered. The filtering range covers high and low temperature environments in all four seasons, the mechanical aging stage of the water meter's entire life cycle, and the full fluid operation conditions of the pipeline network during off-peak / peak / water outage restoration. Invalid and abnormal historical data during historical fault repair and water meter disassembly and calibration are removed, and standardized historical water flow datasets under actual service conditions are retained. This dataset simultaneously includes historical temperature time series data, water meter mechanical operation data, pipeline fluid operation data, and the actual metering error labels of the water meter for the corresponding time period.

[0067] 2. Subsequently, the grey relational analysis method was introduced to calculate the grey relational degree between temperature parameters, water meter mechanical parameters, fluid operating parameters and the overall measurement error of electronic water meter: the fitting similarity between the trend of fluctuation of the three types of parameters and the trend of cumulative measurement error of water meter was statistically analyzed, and the contribution of each type of key parameter to the overall abnormal water flow and measurement deviation of water meter was quantified.

[0068] 3. Combining four typical water supply site conditions—high-rise secondary water supply, low-rise direct supply, old pipe network, and brand-new pipe network—the correlation degree value is dynamically optimized. After normalization, the weights of the first parameter corresponding to the temperature parameter, the second parameter corresponding to the water meter mechanical parameter, and the third parameter corresponding to the fluid condition parameter are output respectively. The sum of the three weights is always 1, accurately representing the actual impact ratio of the three types of interference factors on the water meter flow error in the current water use scenario.

[0069] This invention differs from fixed static weighting schemes by dynamically calculating weights using real, full-cycle historical operating data of water meters. It adapts to the differentiated error distribution of water meters in different installation environments and with different service years. It accurately distinguishes the objective laws that fluid pressure fluctuations and pipeline air accumulation are the dominant errors in most field situations, while high and low temperatures are secondary errors, and the errors from mechanical wear of old water meters gradually increase. This avoids the error assessment distortion caused by fixed weights and ensures that subsequent error corrections closely match the actual operating conditions of the water meters.

[0070] Furthermore, the detailed processing procedure for correcting the water quantity error coefficient based on parameter weights is as follows: This invention relies on the weights of three types of dynamic parameters obtained from the solution to perform weighted adaptive correction processing on the unweighted original first water volume error coefficient, second water volume error coefficient, and third water volume error coefficient, respectively, to achieve working condition adaptation calibration of the original error coefficients. The specific step-by-step correction logic is as follows: 1. Correction of the first water volume error coefficient (temperature dimension): retrieve the weight of the first parameter corresponding to the temperature, multiply the original first water volume error coefficient with the weight of the first parameter, and make a secondary fine adjustment in combination with the current real-time ambient temperature level to eliminate the weight adaptation deviation under extreme high and low temperatures, and obtain the first correction parameter corresponding to the temperature parameter. This parameter is the accurate measurement error of the temperature dimension after fitting the current on-site environment and considering the actual impact ratio.

[0071] 2. Correction of the second water volume error coefficient (mechanical dimension): The weight of the second parameter corresponding to the water meter's mechanical properties is retrieved, and the service life compensation of the second parameter weight is performed by matching the actual service time of the water meter and the aging degree of the movement. Then, it is weighted and calculated with the original second water volume error coefficient. Invalid mechanical error components when the newly installed water meter's mechanical properties are not worn are removed to obtain the second correction parameter corresponding to the water meter's mechanical parameters.

[0072] 3. Correction of the third water volume error coefficient (fluid condition dimension): The weight of the third parameter corresponding to the fluid condition is retrieved, and the weight is dynamically compensated for the condition by combining the daily pipeline water pressure fluctuation level and pipeline air content. The compensated weight is then weighted and fitted with the original third water volume error coefficient to filter out the weak and invalid errors of fluid disturbance under stable water use conditions, and the third correction parameter corresponding to the fluid condition parameter is obtained.

[0073] After weighted correction, this invention unifies the data dimensions and error precision of the three types of correction parameters, ensuring that the three types of correction parameters can be directly integrated to calculate the overall comprehensive metering error of the water meter, and realizing unified quantitative benchmarking of multi-source errors.

[0074] The technical effects of the correction process are as follows: 1. This invention solves the defect of assigning equal weights to the original error coefficients: the original three types of error coefficients assume that the influence of the three types of interference factors is the same, which is inconsistent with the actual operation of pipe networks and water meters. After weighted correction, the true contribution of each error source under different scenarios can be objectively restored, and the accuracy of error quantification is improved by more than 25%.

[0075] 2. This invention achieves adaptive adaptation throughout the entire life cycle of water meters: the weights and correction parameters can be dynamically updated according to the usage time of the water meter, seasonal temperature changes, and changes in pipeline renovation conditions. Newly installed water meters focus on fluid condition error correction, while older water meters simultaneously increase the proportion of mechanical error correction, thus adapting to the metering compensation needs of water meters throughout their entire life cycle from new installation to aging.

[0076] 3. This invention weakens the interference of minor invalid errors: by suppressing minor error terms with extremely low weight, it avoids invalid errors caused by slight temperature fluctuations and slight mechanical gaps from interfering with the overall measurement results, thereby improving the stability and accuracy of subsequent water meter comprehensive error compensation and water flow calibration.

[0077] 4. The final output includes three types of correction parameters—temperature, mechanical, and fluid conditions—that are more reliable and closely match the actual operating conditions on site. This provides a precise and adaptive quantitative basis for subsequent calculation of the total metering error of the water meter by integrating multi-dimensional correction parameters and completing dynamic compensation for the real-time water flow of the water meter.

[0078] Step S50: Adjust the corresponding key parameters of water flow in the electronic water meter according to each of the correction parameters to obtain the adjustment results and the target water flow results.

[0079] This invention uses the first, second, and third correction parameters obtained after weighted correction to match the electronic water meter temperature compensation module, mechanical metering calibration module, and fluid condition adaptation module, respectively. It performs closed-loop adaptive dynamic calibration and parameter compensation adjustment on the real-time temperature parameters, water meter mechanical parameters, and fluid condition parameters collected on-site. It corrects deviations, compares thresholds, and iteratively optimizes each of the three types of parameters until all three types of operating parameters fall within their respective preset standard parameter allowable fluctuation ranges, outputting the final parameter adjustment result. Then, based on the optimal operating parameters after calibration, it re-performs high-precision water flow metering and outputs accurate target water flow results free from multi-source error interference.

[0080] Specifically, the temperature parameter, mechanical parameter, and fluid condition parameter in the electronic water meter are adjusted according to the correction parameters until they are all within the corresponding preset parameter ranges, thus obtaining the adjustment result; the water flow data of the electronic water meter is collected based on the adjustment result to obtain the target water flow result.

[0081] First, the first correction parameter is retrieved as the temperature deviation compensation benchmark value. The real-time monitored field temperature parameter is compared and verified with the preset temperature parameter range (usually 5℃~45℃ standard measurement temperature range) calibrated at the water meter's factory. If the real-time temperature parameter exceeds the standard range or the temperature deviation amplitude exceeds the allowable threshold, the electronic water meter's built-in temperature compensation chip is driven by the first correction parameter to perform layered compensation adjustment: 1. For low-temperature conditions, the flow metering pulse gain coefficient is dynamically increased to offset the small measurement deviation caused by increased water viscosity and slower water flow rate at low temperatures; 2. For high-temperature conditions, the metering pulse acquisition frequency is decreased to compensate for multiple measurement deviations caused by thermal expansion and contraction deformation of the plastic components of the mechanism and circuit temperature drift. The temperature compensation coefficient of the metering circuit is continuously iteratively adjusted, and the adjusted temperature parameter is fed back in real time for secondary verification until the water temperature measurement parameter and the circuit board operating temperature parameter both return to the preset temperature parameter fluctuation range, completing the closed-loop calibration of the temperature dimension.

[0082] This invention achieves dynamic temperature compensation across the entire temperature range of electronic water meters without any blind spots. It solves the problem that the fixed temperature compensation algorithm of traditional water meters cannot adapt to the temperature differences of all seasons and day and night. It completely eliminates the inherent measurement errors caused by circuit temperature drift, water viscosity changes, and thermal deformation of the movement, allowing the water meter to maintain its rated measurement accuracy even in extreme high and low temperature environments.

[0083] Furthermore, the present invention uses the second correction parameter as the basis for mechanical error calibration, and compares the four types of water meter mechanical parameters—real-time impeller pulse frequency, core transmission clearance, check valve opening and closing response time, and water meter installation tilt angle—with the preset standard mechanical parameter threshold range of the water meter one by one.

[0084] To address impeller pulse deviation caused by wear and tear on the mechanism, an electronic pulse compensation algorithm is used to correct the Hall sensor pulse counting benchmark, compensating for pulse loss due to mechanical transmission losses. To address backflow and idling issues caused by poor sealing of the check valve, the built-in electronic anti-backflow calibration logic of the water meter is dynamically activated to counteract invalid metering pulses caused by the reciprocating impact of water flow. To address installation tilt offset deviation, a background algorithm completes mechanical tilt error fitting compensation, eliminating the need for physical adjustment of the water meter's installation position. After multiple rounds of iterative calibration, all mechanical operating parameters are brought back to the preset standard mechanical parameter range, eliminating the original mechanical metering errors caused by the aging of the water meter's hardware and installation deviations.

[0085] This invention eliminates the need for disassembly and repair or replacement of the core hardware. It uses an electronic control algorithm to perform soft calibration of the water meter's mechanical parameters, adapting to the aging and decay patterns of the core throughout the water meter's lifecycle. It effectively suppresses mechanical problems such as impeller wear, check valve failure, and installation tilt that lead to over-counting of water usage while under-counting actual water consumption. This reduces the cost of after-sales maintenance and disassembly for water meters, while extending the service life of the electronic water meter for accurate metering.

[0086] This invention adapts and adjusts the fluid conditions of the pipeline network based on a third correction parameter. It compares real-time pipeline water pressure, water pressure fluctuation amplitude, air content in the pipe, water turbulence, and water flow direction with preset standard stable fluid condition parameter ranges to determine deviations. The specific implementation steps are as follows: 1. For pipeline water pressure fluctuations and water hammer impact conditions, dynamically activate the water meter's built-in adaptive filtering metering mode to shield against instantaneous pressure pulse interference; 2. For pipeline air accumulation and gas-liquid two-phase flow conditions after water supply interruption and restoration, reduce the counting weight of invalid pulses under bubble disturbance and eliminate false water flow pulses generated by bubbles driving impeller rotation; 3. For bidirectional water flow reversal conditions, activate a bidirectional water flow discrimination algorithm to offset duplicate metering values ​​caused by reverse water flow; 4. Continuously adaptively adjust the water meter's fluid metering discrimination threshold until all fluid condition-related operating parameters match the preset standard parameter range, shielding against external metering interference caused by complex pipeline network conditions.

[0087] This invention is suitable for complex on-site conditions such as high-rise secondary water supply, unstable water pressure in old pipe networks, and frequent air accumulation in pipes. It accurately removes invalid and false water volume caused by pipe network disturbances, distinguishes between real water flow and disturbed false water flow, and solves common industry pain points such as electronic water meters running without water and abnormally high water volume from the source of metering.

[0088] Furthermore, after the temperature parameters, water meter mechanical parameters, and fluid condition parameters are all calibrated to the standard preset parameter range and there is no coupling deviation among the three types of parameters, the current optimal metering working parameters are locked, and the electronic water meter flow acquisition module is controlled to re-acquire water flow data within the preset time period with standard sampling frequency and standard pulse counting rules. The original metering logic before error compensation is closed throughout the process. Based on the calibrated pure operating parameters, water flow pulse statistics and cumulative water volume calculation are completed, and finally, a high-precision target water flow result after eliminating temperature error, mechanical error, and fluid condition error is output.

[0089] This invention achieves optimal matching of internal metering parameters and external adaptation parameters of electronic water meters across the entire domain through multi-parameter closed-loop iterative adjustment, completing comprehensive compensation for multi-source coupled metering errors. Compared with the uncalibrated original water flow, the target water flow completely eliminates metering deviations caused by three types of interference factors: environment, hardware, and pipeline network. The metering accuracy is greatly improved, which can truly reflect the actual water consumption of users and meet the actual business needs of water companies for accurate meter reading and fair billing.

[0090] Furthermore, such as Figure 2 As shown, based on the above-mentioned method for calibrating the mechanical error of an electronic water meter, the present invention also provides a mechanical error calibration system for an electronic water meter, wherein the mechanical error calibration system for the electronic water meter includes: The water flow data acquisition module 51 is used to determine multiple key parameters of water flow of the electronic water meter, and to acquire water flow data according to each of the key parameters of water flow to obtain multiple target water flow data. The water flow data processing module 52 is used to perform data preprocessing and feature extraction processing on multiple target water flow data to obtain multiple target water flow features; The water volume error coefficient calculation module 53 is used to calculate the predicted water flow corresponding to each of the target water flow characteristics, and to calculate multiple water volume error coefficients based on the multiple predicted water flow and the actual water flow. The water volume error coefficient correction module 54 is used to obtain the parameter weight corresponding to each of the key parameters of water flow, and to correct each of the water volume error coefficients according to the parameter weights to obtain multiple correction parameters. The water flow result output module 55 is used to adjust the corresponding key parameters of water flow in the electronic water meter according to each of the correction parameters, so as to obtain the adjustment result and the target water flow result.

[0091] Furthermore, such as Figure 3 As shown, based on the above-mentioned mechanical error calibration method and system for electronic water meters, the present invention also provides a terminal, which includes a processor 10, a memory 20 and a display 30. Figure 3Only some of the terminal components are shown; however, it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0092] In some embodiments, the memory 20 may be an internal storage unit of the terminal, such as a hard disk or memory. In other embodiments, the memory 20 may be an external storage device of the terminal, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc. Further, the memory 20 may include both internal and external storage devices. The memory 20 is used to store application software and various types of data installed on the terminal, such as the program code installed on the terminal. The memory 20 can also be used to temporarily store data that has been output or will be output. In one embodiment, the memory 20 stores a mechanical error calibration program 40 for an electronic water meter, which can be executed by the processor 10 to implement the mechanical error calibration method for the electronic water meter in this application.

[0093] In some embodiments, the processor 10 may be a central processing unit (CPU), a microprocessor, or other data processing chip, used to run program code stored in the memory 20 or process data, such as executing the mechanical error calibration method of the electronic water meter.

[0094] In some embodiments, the display 30 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. The display 30 is used to display information on the terminal and to display a visual user interface.

[0095] In one embodiment, when the processor 10 executes the mechanical error calibration program 40 of the electronic water meter in the memory 20, it implements the steps of the mechanical error calibration method of the electronic water meter as described above.

[0096] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a mechanical error calibration program for an electronic water meter, and the mechanical error calibration program for the electronic water meter, when executed by a processor, implements the steps of the mechanical error calibration method for the electronic water meter as described above.

[0097] In summary, this invention provides a method, system, terminal, and storage medium for calibrating the mechanical error of an electronic water meter. The method includes: determining multiple key parameters of water flow in the electronic water meter, and collecting water flow data based on each key parameter to obtain multiple target water flow data; performing data preprocessing and feature extraction on the multiple target water flow data to obtain multiple target water flow features; calculating the predicted water flow corresponding to each target water flow feature, and calculating multiple water flow error coefficients based on the multiple predicted water flow and actual water flow; obtaining the parameter weights corresponding to each key parameter of water flow, and correcting each water flow error coefficient based on the parameter weights to obtain multiple correction parameters; adjusting the corresponding key parameter of water flow in the electronic water meter based on each correction parameter to obtain the adjustment result and the target water flow result. This invention calculates the water flow error coefficients by predicting the water flow and the actual water flow, and then adjusts the parameters of the electronic water meter based on the water flow error coefficients, which can effectively improve the accuracy of the water flow result.

[0098] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal that includes that element.

[0099] Of course, those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.). The program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The computer-readable storage medium can be a memory, magnetic disk, optical disk, etc.

[0100] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A mechanical error calibration method of an electronic water meter, characterized by, The mechanical error calibration method for the electronic water meter includes: Multiple key parameters of water flow in the electronic water meter are determined, and water flow data is collected based on each of the key parameters to obtain multiple target water flow data. Data preprocessing and feature extraction are performed on multiple target water flow data to obtain multiple target water flow features; Calculate the predicted flow rate corresponding to each of the target water flow characteristics, and calculate multiple water flow error coefficients based on the multiple predicted flow rates and the actual flow rates; Obtain the parameter weight corresponding to each of the key parameters of water flow, and perform correction processing on each of the water flow error coefficients according to the parameter weights to obtain multiple correction parameters; The corresponding key parameters of water flow in the electronic water meter are adjusted according to each of the correction parameters to obtain the adjustment results and the target water flow results.

2. The mechanical error calibration method of an electronic water meter according to claim 1, characterized in that, The process involves determining multiple key parameters of water flow from the electronic water meter, and collecting water flow data based on each of these key parameters to obtain multiple target water flow data. Specifically, this includes: Several key parameters affecting the water flow of electronic water meters were identified, including temperature parameters, water meter mechanical parameters, and fluid condition parameters. Water flow data is collected within a preset time period based on the temperature parameter, the water meter mechanical parameter, and the fluid condition parameter, respectively, to obtain multiple target water flow data.

3. The mechanical error calibration method for electronic water meters according to claim 1, characterized in that, The process of preprocessing and feature extraction of multiple target water flow data to obtain multiple target water flow features specifically includes: Data filtering, integrity detection, and missing value imputation are performed on multiple target water flow data to obtain multiple preprocessed data; A preset deep learning model is used to extract features from multiple preprocessed data to obtain multiple target water flow features.

4. The mechanical error calibration method for electronic water meters according to claim 1, characterized in that, The target water flow characteristics include temperature characteristics, water meter mechanical characteristics, and fluid operating condition characteristics; The water volume error coefficient includes a first water volume error coefficient, a second water volume error coefficient, and a third water volume error coefficient; The calculation of the predicted flow rate corresponding to each of the target water flow characteristics, and the calculation of multiple water flow error coefficients based on the multiple predicted flow rates and the actual flow rates, specifically includes: Determine the actual water flow of the electronic water meter; Calculate the first predicted water flow corresponding to the temperature characteristic, and calculate the difference between the first predicted water flow and the actual water flow to obtain the first water flow error coefficient; Calculate the second predicted water flow corresponding to the mechanical characteristics of the water meter, and calculate the difference between the second predicted water flow and the actual water flow to obtain the second water flow error coefficient; Calculate the third predicted water flow rate corresponding to the fluid operating condition characteristics, and calculate the difference between the third predicted water flow rate and the actual water flow rate to obtain the third water flow error coefficient.

5. The mechanical error calibration method for electronic water meters according to claim 4, characterized in that, The step involves obtaining the parameter weights corresponding to each of the key parameters of water flow, and then correcting the error coefficient of each water flow based on the parameter weights to obtain multiple correction parameters, specifically including: Obtain historical water flow data from the electronic water meter, and obtain the parameter weight corresponding to each of the key parameters of water flow based on the historical water flow data; Based on the parameter weights, the first water volume error coefficient, the second water volume error coefficient, and the third water volume error coefficient are corrected to obtain the correction parameters corresponding to each key water flow parameter.

6. The mechanical error calibration method for electronic water meters according to claim 2, characterized in that, The step of adjusting the key parameters of water flow in the electronic water meter according to each of the correction parameters to obtain the adjustment result and the target water flow result specifically includes: The temperature parameter, mechanical parameter, and fluid condition parameter in the electronic water meter are adjusted according to the correction parameters until the temperature parameter, mechanical parameter, and fluid condition parameter are all within the corresponding preset parameter range, thus obtaining the adjustment result. Based on the adjustment results, the water flow data of the electronic water meter is collected to obtain the target water flow result.

7. The mechanical error calibration method for electronic water meters according to claim 3, characterized in that, The data filtering process includes threshold filtering, sliding filtering, and Kalman filtering.

8. A mechanical error calibration system for an electronic water meter, characterized in that, The mechanical error calibration system of the electronic water meter includes: The water flow data acquisition module is used to determine multiple key parameters of water flow of the electronic water meter, and to acquire water flow data based on each of the key parameters of water flow to obtain multiple target water flow data. The water flow data processing module is used to perform data preprocessing and feature extraction on multiple target water flow data to obtain multiple target water flow features; The water volume error coefficient calculation module is used to calculate the predicted water flow corresponding to each of the target water flow characteristics, and to calculate multiple water volume error coefficients based on the multiple predicted water flow and the actual water flow. The water volume error coefficient correction module is used to obtain the parameter weights corresponding to each of the key parameters of water flow, and to correct each of the water volume error coefficients according to the parameter weights to obtain multiple correction parameters. The water flow result output module is used to adjust the corresponding key parameters of water flow in the electronic water meter according to each of the correction parameters, so as to obtain the adjustment result and the target water flow result.

9. A terminal, characterized in that, The terminal includes: a memory, a processor, and a mechanical error calibration program for an electronic water meter stored in the memory and executable on the processor. When the mechanical error calibration program for the electronic water meter is executed by the processor, it implements the steps of the mechanical error calibration method for an electronic water meter as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a mechanical error calibration program for an electronic water meter, which, when executed by a processor, implements the steps of the mechanical error calibration method for an electronic water meter as described in any one of claims 1-7.

Images