Intelligent commercial electronic scale parameter management system integrated with data processing module

Abstract

The invention relates to the technical field of electronic weighing, and discloses an intelligent commercial electronic scale parameter management system integrated with a data processing module, which comprises a sensor, a conversion module, a storage module and a processing module. When monitoring a loading state, the processing module calculates a product of a load amplitude and a sampling time interval and accumulates the product to an integral register to record a stress historical integral value; when an unloading event is monitored, correcting the basic viscosity coefficient according to a linear function relationship between a stress historical integral value and a stress sensitive factor, and obtaining a dynamic viscosity coefficient and a dynamic relaxation time constant; calculating a prediction residual error amount which exponentially decays along with time based on the dynamic relaxation time constant; and finally, subtracting the predicted residual error amount from an original sampling value to obtain a net weight value. According to the invention, by quantifying the creep characteristics of the stress history dynamic compensation sensor, the problems of return-to-zero lag and zero drift are solved, and the metering precision and stability of the electronic scale under continuous weighing operation are improved.

Description

Technical Field

[0001] This invention relates to the field of electronic weighing technology, specifically to an intelligent commercial electronic scale parameter management system with an integrated data processing module. Background Technology

[0002] Commercial electronic scales rely on resistance strain gauge sensors to convert physical pressure into electrical signals for measurement. The core component of these sensors is a metallic elastomer, which undergoes elastic deformation under external force, and the deformation is proportional to the load mass. In continuous weighing operations, a microprocessor acquires weight data by collecting changes in the electrical signal output from the sensor bridge and performing analog-to-digital conversion. This data processing relies on a preset mathematical model to map the voltage signal into a mass reading.

[0003] However, the inherent viscoelastic properties of the metallic elastomer materials in sensors lead to hysteresis and creep during loading and unloading cycles. Existing signal processing techniques employ fixed filtering parameters or static compensation models, neglecting the dynamic impact of the cumulative effect of stress history on the material's rheological properties. Specifically, after a static weighing task with a large load and long duration, the residual deformation amplitude and recovery rate during the unloading phase differ significantly from those after a weighing task with a small load and short duration. Traditional algorithms cannot distinguish between these different stress history states, resulting in discrepancies between the theoretical calculation model and the actual physical state of the material.

[0004] This parameter mismatch causes the electronic scale to experience slow zero-return or non-linear zero-point drift during the unloading phase, specifically manifested as the displayed value failing to immediately stabilize at absolute zero. The resulting residual error is directly added to the next weighing measurement, reducing the linearity and repeatability of continuous weighing data. Therefore, this invention proposes an intelligent commercial electronic scale parameter management system with an integrated data processing module to address the shortcomings of existing technologies. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent commercial electronic scale parameter management system with an integrated data processing module. This system solves the problems that existing commercial electronic scales, after experiencing weighing operations with different load intensities and holding times, suffer from issues such as delayed zeroing of readings during the unloading phase, difficulty in eliminating residual errors, and nonlinear drift of the system zero point, which in turn affect the accuracy of subsequent measurements due to the viscoelastic creep characteristics of the sensor's elastomer material.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solution: an intelligent commercial electronic scale parameter management system with an integrated data processing module, including a conversion module connected to a sensor for converting physical pressure into an analog voltage signal and outputting digital raw sampled values; The storage module stores the basic elastic modulus, basic viscosity coefficient, and stress sensitivity factor. The processing module is configured to: When the sensor is detected to be under load, the product of the load amplitude determined based on the original sampled value and the sampling time interval is calculated, and the calculation result is updated to the integration register to record the stress history integration value. When an unloading event is detected by the sensor, the basic viscosity coefficient is corrected according to the linear function relationship between the stress history integral value and the stress sensitivity factor, and the dynamic viscosity coefficient corresponding to the current weighing is calculated. The dynamic relaxation time constant is determined by the ratio of the dynamic viscosity coefficient to the basic elastic modulus, and the predicted residual error, which decays exponentially with time, is calculated based on the dynamic relaxation time constant. The corrected net weight value is obtained by subtracting the predicted residual error from the original sampled value and outputting it as the final weighing result.

[0007] Preferably, the process by which the processing module determines the load amplitude and updates the stress history integral value is as follows: reading the system zero-point value maintained in the storage module, subtracting the system zero-point value from the original sampled value to obtain the effective load value, and using the effective load value as the load amplitude; comparing the effective load value with a preset loading threshold, and if the effective load value is continuously greater than the loading threshold, determining that the sensor is in the loading state; using a trapezoidal numerical integration algorithm, calculating the single-step stress increment using the effective load value at the current sampling time, the effective load value at the previous sampling time, and the actual sampling time interval, and accumulating the single-step stress increment to the integration register.

[0008] Preferably, the process by which the processing module corrects the basic viscosity coefficient is as follows: calculating the product of the stress history integral value and the stress sensitivity factor, and adding the product to a constant 1 to obtain a dynamic correction ratio; multiplying the basic viscosity coefficient by the dynamic correction ratio to obtain the dynamic viscosity coefficient, wherein the dynamic viscosity coefficient is used to characterize the macroscopic viscous resistance of the sensor material after experiencing stress history.

[0009] Preferably, the process by which the processing module calculates the predicted residual error is as follows: recording the unloading time when the unloading event occurs, and calculating the initial residual amplitude at the moment of unloading using the stress history integral value and the dynamic relaxation time constant; obtaining the absolute elapsed time of the current time relative to the unloading time, and determining the predicted residual error at the current time according to the natural exponential decay law based on the initial residual amplitude, the absolute elapsed time, and the dynamic relaxation time constant; and clearing the integral register when the processing module detects the loading state again.

[0010] Preferably, the processing module uses a recursive algorithm to implement the process of determining the predicted residual error according to the natural exponential decay law: pre-calculate the single-step decay factor based on the dynamic relaxation time constant and the system sampling period; multiply the predicted residual error calculated at the previous sampling time by the single-step decay factor to obtain the predicted residual error at the current time.

[0011] Preferably, the processing module is further configured to perform integral saturation protection: a preset integral upper limit value is set, which is determined based on the product of the maximum range of the electronic scale and the time required for the sensor material to reach creep saturation; after each update of the integral register, the processing module determines whether the historical stress integral value exceeds the integral upper limit value; if the historical stress integral value exceeds the integral upper limit value, the value of the integral register is forcibly set to the integral upper limit value.

[0012] Preferably, the processing module monitors the unloading event by: calculating the rate of change of the original sampled value within a preset sliding window, and comparing the rate of change with a preset negative unloading slope threshold; when the rate of change is less than the unloading slope threshold and the duration exceeds the anti-shake cycle, it is determined that the sensor has experienced the unloading event, and write operations to the integration register are prohibited.

[0013] Preferably, after outputting the net weight value, the processing module further performs the following operations: comparing the net weight value with a preset zero-point dead zone threshold; if the net weight value is less than or equal to the zero-point dead zone threshold, then forcibly assigning the net weight value to zero; using the forcibly assigned net weight value as an input signal to run an automatic zero-point tracking algorithm, which is used to correct the system's zero-point drift in the event of residual deformation at the sensor physical level.

[0014] Preferably, the basic elastic modulus and the basic viscosity coefficient are obtained through the following calibration method: a short-time standard step excitation is applied to the sensor under constant temperature conditions, and the load magnitude of the standard step excitation is recorded; at the instant the standard step excitation is applied, the instantaneous response amplitude of the sensor is acquired, and the ratio of the load magnitude to the instantaneous response amplitude is calculated as the basic elastic modulus; after the short-time standard step excitation is removed, free decay response data is acquired, the residual response signal in the free decay response data is extracted, and a logarithmic linear regression analysis is performed on the relationship between the residual response signal and time; the basic time constant is determined according to the slope of the regression line, and the basic time constant is determined as the basic viscosity coefficient.

[0015] Preferably, the stress sensitivity factor is obtained through the following calibration method: performing multiple cyclic tests on the sensor, applying the same standard load but setting different loading holding times in each cyclic test to form a variable-duration gradient loading test sequence; for each cyclic test, recording the cumulative integral value at the moment of unloading, and calculating the observation time constant after unloading; constructing a dataset with the cumulative integral value as the independent variable and the rate of change of the observation time constant relative to the basic time constant as the dependent variable; performing least squares fitting through the origin on the dataset, and determining the slope obtained by fitting as the stress sensitivity factor.

[0016] This invention provides an intelligent commercial electronic scale parameter management system with an integrated data processing module. It has the following beneficial effects: 1. This invention, by configuring an integral register in the processing module, calculates the product of the load amplitude and the sampling time interval in real time and accumulates it to form a stress history integral value, thereby realizing the quantitative tracking of the historical stress state of the sensor material. This invention uses the linear function relationship between the stress history integral value and the stress sensitivity factor to dynamically correct the basic viscosity coefficient. This method enables the intelligent commercial electronic scale parameter management system with integrated data processing module to distinguish the different degrees of creep effects caused by short-term light loads and long-term heavy loads on the sensor, ensuring that the current physical rheological state of the sensor material can be matched when calculating the dynamic relaxation time constant, thereby improving the measurement accuracy of commercial electronic scales under frequent and continuous weighing conditions.

[0017] 2. This invention utilizes a natural exponential decay model based on a dynamic relaxation time constant to calculate and predict residual error in real time. This predicted residual error is then subtracted from the original sampled value during digital signal processing. This method solves the mechanical zero-return hysteresis problem caused by the viscoelastic effect of elastomer materials. By pre-eliminating the influence of residual deformation that has not yet fully recovered at the physical level at the data level, this invention eliminates the display tailing phenomenon at the moment of unloading, ensuring that the displayed value can quickly and accurately return to zero. Combined with an integral saturation protection mechanism, this invention prevents algorithm divergence caused by extreme usage conditions, suppresses the accumulation of system zero-point drift, and improves the long-term operational stability of the equipment.

[0018] 3. This invention proposes a step-by-step calibration strategy that independently obtains the basic elastic modulus and basic viscosity coefficient through short-time standard step excitation, and obtains the stress sensitivity factor through a variable-duration gradient loading test sequence. This strategy avoids the over-reliance on data samples by purely mathematical fitting methods. By accurately measuring the basic elastic modulus, the basic viscosity coefficient, and the stress sensitivity factor, this invention can adapt to individual differences in sensors from different batches and with different materials, reducing the system's requirements for sensor hardware consistency, thereby reducing production and debugging costs and enhancing the system's versatility. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the system architecture of the present invention; Figure 2 This is a schematic diagram of the overall system logic flow of the present invention; Figure 3 This is a schematic diagram of the full load timing of the present invention; Figure 4 This is a schematic diagram illustrating the dynamic correction of model parameters in this invention; Figure 5 This is a schematic diagram comparing the signals during unloading and rapid loading in this invention.

[0020] Among them, 10 is the sensor; 20 is the conversion module; 30 is the processing module; 40 is the storage module; and 31 is the integration register. Detailed Implementation

[0021] The technical solutions in 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.

[0022] See attached document Figure 1This invention provides an intelligent commercial electronic scale parameter management system with an integrated data processing module. The system mainly includes: a sensor 10, a conversion module 20, a processing module 30, and a storage module 40. The sensor 10 is physically connected to the electronic scale's support mechanism and is used to convert the physical pressure received by the support mechanism into an analog voltage signal. The conversion module 20 is electrically connected to the sensor 10 and is configured to perform sampling at a preset period. The analog voltage signal output by sensor 10 is discretized and sampled, and then converted into a digital form of raw sampled value. The processing module 30 is communicatively connected to the conversion module 20 and the storage module 40, respectively, and is used to receive the raw sampled values. It also performs parameter management and dynamic compensation calculations. The storage module 40 is used to store various parameters and temporary data required for system operation.

[0023] The storage module 40 has a pre-set basic parameter area, which stores the basic elastic modulus characterizing the material properties of the sensor 10. Basic viscosity coefficient and stress-sensitive factors The processing module 30 is internally configured with an integration register 31, which is used to store the cumulative product of load amplitude and time in real time during load loading.

[0024] See attached document Figure 2 The present invention provides an intelligent commercial electronic scale parameter management system with an integrated data processing module for executing an intelligent commercial electronic scale parameter management method with an integrated data processing module, the method comprising the following steps: System initialization and parameter reading: After the system is powered on, the processing module 30 reads the basic elastic modulus from the storage module 40. Basic viscosity coefficient and stress-sensitive factors The integration register 31 is then cleared. The conversion module 20 continuously transmits the raw sampled values ​​to the processing module 30. .

[0025] Execution of load monitoring and stress history integration: Processing module 30 monitors raw sampled values ​​in real time. When the original sampled value When the load exceeds a preset loading threshold, the processing module 30 determines that the sensor 10 is in a loaded state. The preset sampling period during the loaded state... Inside, processing module 30 performs discrete integration to calculate the current time step. Stress history integral value And update to integration register 31. Stress history integral values The calculation formula is as follows: ; in, This represents the accumulated value of the integration register 31 at the previous sampling time. This represents the original sampled value at the current sampling moment. The integration register 31 records the historical cumulative force of sensor 10 during the current weighing transaction.

[0026] Unloading determination and model parameter correction: Processing module 30 calculates the original sampled values. Rate of change. When the original sample value When the rate of change is less than the preset unloading slope threshold, the processing module 30 determines that the sensor 10 has experienced an unloading event. The processing module 30 locks the current value of the integration register 31 as the final integration value. Using the final integral value For the basic viscosity coefficient Make corrections and calculate the dynamic viscosity coefficient corresponding to the current weighing transaction. Dynamic viscosity coefficient The calculation formula is as follows: ; Processing module 30 based on dynamic viscosity coefficient With basic elastic modulus Determine the dynamic relaxation time constant used to describe the recovery rate of residual deformation after unloading. The calculation relationship is as follows .

[0027] Perform residual error prediction and compensation output: After the unloading event occurs, the processing module 30 utilizes the dynamic relaxation time constant. Calculate the current time Relative to unloading time Prediction residual error Predicted residual error The calculation formula is as follows: ; in, This is a preset proportional coefficient. This refers to the unloading moment when the unloading event occurs. Processing module 30 will convert the raw sampled values ​​input from conversion module 20. Subtract the prediction residual error The corrected net weight value is obtained. and net weight value This serves as the final weighing result output of the system. When the processing module 30 detects a loading event again, it clears the integration register 31 and returns to the execution of the loading monitoring and stress history integration steps.

[0028] See attached document Figure 2The specific process of loading monitoring and stress history integration performed by processing module 30 includes the following sub-steps: The core logic of this step lies in constructing a stress memory model of the sensor's elastomer. Because the elastomer material of the resistance strain gauge load cell has viscoelastic characteristics, its deformation recovery process depends not only on the magnitude of the currently unloaded weight but also on the duration of that weight's residence on the sensor. Prolonged loading leads to deeper slippage and orientation of the polymer chain segments, resulting in greater mechanical hysteresis. Therefore, this embodiment quantifies this stress history effect by integrating the effective load over time.

[0029] Acquiring payload data: Processing module 30 receives the raw sampled values ​​transmitted by conversion module 20. Processing module 30 reads the system zero-point value maintained in real time from storage module 40. and will Subtract the system zero point value To obtain the payload value at the current moment. The system zero point value It is not fixed, but a reference value that is dynamically updated when the sensor is idle through a zero-point tracking algorithm, which is used to eliminate static errors caused by temperature drift and dust accumulation on the weighing pan.

[0030] Determine the validity of the loading status: Processing module 30 will determine the effective payload value. With preset loading threshold Compare. Load threshold. The value range is set as the calibration scale division value of the electronic scale. The value is 3 to 10 times that of the sensor and its peripheral circuitry, with the specific value determined based on the background noise level of the sensor and its peripheral circuitry to ensure that the integration process is not affected by minor environmental vibrations or airflow interference. When the effective payload value... Continuously exceeding the loading threshold When the processing module 30 determines that the sensor 10 is in an effective loading state, it indicates that the sensor is undergoing substantial viscoelastic deformation accumulation, and at this time, updating the integration register 31 is allowed; when the effective load value Less than or equal to the loading threshold When the sensor 10 is in an idle or invalid perturbation state, the processing module 30 determines that the sensor 10 is in an idle or invalid perturbation state. At this time, the current value of the integration register 31 is kept unchanged to prevent the accumulation of invalid noise signals from causing the model parameters to drift incorrectly.

[0031] Perform discrete-time integration: in each preset sampling period when the system is determined to be in a valid loading state. Internally, processing module 30 uses the rectangular integration method to accumulate stress history. Processing module 30 reads the accumulated value stored in integration register 31 at the previous sampling time. The original sampled value at the current sampling time With preset sampling period The product and cumulative value By adding them together, the stress history integral value at the current moment can be calculated. .

[0032] Execute integral saturation protection: Processing module 30 updates the stress history integral value each time. Then, the calculation results are compared with the preset upper limit of integration. Compare the scores. Maximum score. It is the saturation constant determined based on the physical limits of the sensor material, and the calculation formula is: ; in, This is the maximum weighing range of the electronic scale. The empirical time required for the sensor material to reach creep saturation is typically taken as 1800 to 3600 seconds.

[0033] If the calculated stress history integral value Exceeding the maximum points limit The processing module 30 forcibly sets the value of the integration register 31 to the upper limit of integration. The technical purpose of setting this upper limit is to simulate the saturation characteristics of viscous deformation of materials, while preventing numerical overflow caused by abnormally long periods (such as several days) of heavy object stacking, and ensuring that the model parameters are still within the effective convergence range after the system resumes normal weighing.

[0034] Real-time update of the integral register: The processing module 30 updates the historical stress integral value after performing discrete-time integration and verification by performing integral saturation protection. Write to integration register 31, overwriting the old value. This integration value will be maintained or accumulated until it is locked and invoked in a subsequent step when an unload event is detected.

[0035] See attached document Figure 2 Regarding the discrete-time integration operation mentioned in the discrete-time integration operation step, processing module 30 executes the following specific algorithm processing steps: The technical objective of this step is to faithfully reconstruct the energy accumulation history during the sensor's force application process. Due to the dynamic viscosity coefficient in the subsequent model... stress history integral value The integral exhibits a linear sensitivity; if there is a significant truncation error or accumulated error due to clock jitter during the integration process, the error will be amplified, leading to the failure of the final residual prediction. Therefore, this embodiment employs a trapezoidal numerical integration algorithm with variable step size to adapt to the nonlinear loading rate that may occur during the placement of goods on commercial electronic scales.

[0036] Time base acquisition and dynamic step size calculation: Processing module 30 receives the current time... When obtaining the original sampled value, the current count value of the microcontroller's internal hardware timer is synchronously read and recorded as the current timestamp. Processing module 30 reads the timestamp saved from the previous sampling time. Processing module 30 operates according to the operating frequency of the hardware timer. Calculate the actual time interval between the current sampling period and the previous sampling period. The actual time interval is calculated according to the following formula: ; in, The actual time interval is measured in seconds (s). The clock frequency of the hardware timer (e.g., 1MHz or higher). If the current time... If the loading status is determined at the beginning of the frame (i.e., immediately after the loading event is triggered), then set... Preset sampling period for the system And set the effective payload value at the previous moment. The value is 0. This step captures the actual sampling interval through a hardware counter, eliminating sampling jitter errors caused by software interrupt response delays or multi-task scheduling.

[0037] Performing trapezoidal approximate integration: To reduce truncation errors during discretization, especially during the stage of rapid load changes caused by users quickly placing goods, processing module 30 uses the trapezoidal integration rule for numerical accumulation. Processing module 30 calls the effective load value from the previous moment stored in storage module 40. Combined with the current payload value and actual time interval Update the stress history integral value. The calculation formula is as follows: ; in, This is the updated stress history integral value at the current moment; This is the cumulative value stored at the previous sampling time; in this formula, the term... This represents the average effective load force within the current sampling interval. Compared to the traditional rectangular integral method, this algorithm can more accurately approximate the area under the stress-time curve in the continuous time domain.

[0038] Accuracy control and status update: based on stress history integral values The increment monotonically increases over time, while the increment value in a single sampling period is relatively small. To prevent precision loss due to large numbers swallowing small numbers in floating-point operations, the processing module 30 defines... The calculation uses either double-precision floating-point format or 64-bit long integer fixed-point format conforming to the IEEE 754 standard (Binary Floating-Point Arithmetic Standard). After the calculation is completed, the processing module 30 will... Assign to , will the current Assign to And will update Stored in integration register 31 as the historical state input parameter for the next calculation cycle.

[0039] See attached document Figure 2 Regarding the process of correcting model parameters using the final integral value mentioned in the unloading determination and model parameter correction steps, the processing module 30 executes the following specific parameter calculation and mapping sub-steps: The technical essence of this step lies in constructing a digital viscoelastic state machine. In the classic Kelvin-Voigt rheological model, the material is equivalent to a parallel structure of an ideal spring and a damper. Traditional methods typically treat the damping coefficient as a constant, which cannot explain the nonlinear phenomenon of slowing down the return to zero under heavy pressure over long periods. This embodiment introduces a variable parameter mechanism to mathematically map the physical phenomenon of material hardening or increased viscosity as a dynamic drift of the damping coefficient with the stress history integral value.

[0040] Reading and parsing basic physical parameters: Upon determining that an unloading event has occurred at sensor 10, the final integral value is locked. Then, the processing module 30 accesses the basic parameter area of ​​the storage module 40 and reads the pre-calibrated basic elastic modulus. Basic viscosity coefficient and stress-sensitive factors Among them, the basic elastic modulus With basic viscosity coefficient The stress sensitivity factor is a reference parameter obtained by applying a standard step load to the sensor for a short time (e.g., within 1 second) and observing its transient response curve. It represents the initial physical characteristics of the sensor under conditions without fatigue accumulation. It is an empirical coefficient characterizing the sensitivity of the rheological properties of sensor materials to stress history.

[0041] Calculation of dynamic viscosity coefficient: Processing module 30 uses the final integral value For the basic viscosity coefficient Make adjustments to generate a dynamic viscosity coefficient specific to the current weighing transaction. Based on the physical properties of polymer materials, as the stress duration increases, the untangling and slippage of molecular chain segments within the material become more difficult, resulting in an increase in macroscopic viscous resistance.

[0042] Generate dynamic relaxation time constant: Processing module 30 calculates the dynamic viscosity coefficient... The basic viscosity coefficient of the reading Calculate the dynamic relaxation time constant used to describe the residual error decay rate. The processing module 30 will calculate the dynamic relaxation time constant. It is temporarily stored in a register for use in subsequent residual error prediction steps.

[0043] See attached document Figure 2 For the unloading determination and model parameter correction steps and the subsequent transition phase, the processing module 30 executes the following specific timing control sub-steps: Feature capture of the unloading event: This step aims to accurately distinguish between the user's unloading action and instantaneous negative fluctuations caused by environmental vibrations. Processing module 30 calculates the raw sampled values ​​within each sampling period. The real-time rate of change. Processing module 30 uses a sliding window difference algorithm to calculate the current time. Compared to the previous Average slope at time 1 . The calculation formula is as follows: ; in, The length of the sliding window at any given time, with a value ranging from 3 to 5; The preset sampling period is used. The processing module 30 will calculate the... Compared with the preset uninstallation threshold Compare them. The value is based on the full-scale value of the electronic scale. OK, usually set to -0.5 / s to -2.0 Between / s. The preset unloading threshold ensures that the system only recognizes a valid unloading when the load decreases at a significant and sustained rate. When Furthermore, when the number of consecutive frames exceeds the anti-shake period (e.g., 3 consecutive frames), the processing module 30 determines that the current time is the unloading start time and marks the index of this time as the unloading time. .

[0044] Execution state locking and integral truncation: Once the unloading time is established The processing module 30 immediately disables write operations to the stress history integration register 31. At this time, the value stored in the integration register 31 is fixed as the final integration value. This logical locking mechanism ensures that the integral value used as an input variable is a definite scalar reflecting the complete loading history during subsequent model parameter calculations, avoiding interference from signal oscillations during the unloading process. The processing module 30 also records the timestamp at this point, serving as the zero point for subsequent residual error decay calculations.

[0045] Instantaneous solution of model parameters: at the unloading time During the subsequent parameter calculation cycle, the processing module 30 utilizes the locked final integral value. The dynamic viscosity coefficient specific to the current weighing transaction is calculated based on the parameter coupling formula. and dynamic relaxation time constant These two parameters characterize the specific rheological state reached by the sensor during the immediate stress process. Once calculated, they are loaded into the buffer of the compensation operation and remain constant in the subsequent residual recovery phase until the next new effective loading event is triggered.

[0046] Initializing the residual error amplitude: To establish an accurate compensation benchmark, processing module 30 needs to estimate the total amount of unrecovered deformation accumulated inside the sensor at the moment of unloading. Based on the principle of viscous rheology, the total viscous deformation is directly proportional to the applied stress impulse and inversely proportional to the current viscosity of the material. Processing module 30 calculates the initial amplitude of the residual error using the following formula. : ; in, This is a proportional correction factor, measured in kg / (N·s), used to convert the theoretical deformation calculated by the model into an error value in the weight domain. The specific value is determined during the system calibration phase by comparing the actual residual amount with the calculated value. This formula indicates that although heavy, long-duration loading will lead to a lower final integral value... Increased, but due to the dynamic relaxation time constant The denominator also increases accordingly, and the system can accurately calculate the actual physical hysteresis at the moment of unloading based on the nonlinear proportional relationship between the two, thus preventing over-compensation or under-compensation.

[0047] Execution register reset and state machine transition: After completing the above parameter calculations and initial amplitude settings, processing module 30 switches the system state from monitoring integration state to compensation output state. Processing module 30 clears the contents of integration register 31 and releases its locked state to prepare for a new integration accumulation for the next independent weighing transaction. At this time, the dynamic relaxation time constant calculated by the instantaneous solution step of the execution model parameters is executed. The initial magnitude calculated in the step of initializing the residual error magnitude As a residual attribute of the current unloading event, it is retained and continuously drives the subsequent residual error calculation, realizing the logical decoupling between the current physical zeroing process and the next business data collection.

[0048] See attached document Figure 2 Regarding the process of using the model to track the residual error in real time, as mentioned in the residual error prediction and compensation output step, the processing module 30 performs the following specific prediction calculation sub-steps at each sampling time after the unloading event occurs: Calculate the decay time increment: Processing module 30 at the current moment (in ), read the current count value of the system hardware timer Processing module 30 retrieves the unloading timestamp locked in step S350. And combined with hardware timer frequency Calculate the absolute elapsed time from the start of the self-unloading action to the present. . The calculation formula is as follows: ; in, This step calculates the absolute elapsed time, in seconds (s), from the start of the self-unloading action to the present. It uses the absolute count difference of the hardware clock as a time base, rather than relying on the number of software loops, eliminating accumulated time errors caused by interrupt delays or task scheduling, and ensuring strict synchronization of the decay curves on the time axis.

[0049] Perform exponential decay calculation: Processing module 30 uses the determined initial amplitude and dynamic relaxation time constant Calculate the prediction residual error at the current time. According to the physical definition of the Kelvin-Voigt viscoelastic model, when the external load is removed, the parallel dampers will hinder the instantaneous recovery of the elastic spring. Its deformation rebound process follows the general solution form of the first-order homogeneous linear differential equation, that is, it exhibits a natural exponential decay law.

[0050] Based on the processor's computing power, the processing module 30 selects one of the following two algorithms to perform the calculation: Method 1 (High-precision mode): Processing module 30 directly calls the exponential function in the math library, using the current absolute elapsed time. Perform the calculation: ; This method does not have accumulated calculation errors and is suitable for processors with strong computing performance (such as those with FPU units), and can accurately reproduce the theoretical residuals at any time.

[0051] Method 2 (Low-Power Recursive Mode): To reduce time-consuming exponential function calls, processing module 30 employs a discrete recursive algorithm. During the unloading and initialization phase, processing module 30 pre-calculates the single-step attenuation factor. : ; in, The nominal sampling period of the system (e.g., 0.01s).

[0052] Every subsequent current moment The prediction residual error is obtained through the following recursive formula: ; in, This is the calculation result from the previous moment; in time, Values This algorithm transforms complex exponential operations into a single floating-point multiplication, significantly reducing CPU load and making it suitable for cost- and power-sensitive commercial electronic scale hardware platforms.

[0053] Convergence determination and truncation: Processing module 30 calculates the predicted residual error. With the preset residual zeroing threshold Compare. Residual zeroing threshold. Based on the minimum display scale value It is determined that its value range is... to (For example, for the scale value) electronic scale, (Set to 0.1g).

[0054] when When, the processing module 30 retains the calculated value as a valid basis for compensation; when At that time, processing module 30 determines that the mechanical hysteresis of the system has decayed to an unobservable micro-level. In order to prevent the long-tail characteristic of the model, which mathematically approaches zero but never reaches zero, from interfering with the zero-point tracking logic, processing module 30 forcibly... The value is set to 0, and subsequent decay calculations are stopped until the next new load event is triggered.

[0055] See attached document Figure 2 Regarding the data compensation and output stage mentioned in the residual error prediction and compensation output step, the processing module 30 performs the following specific synthesis sub-steps: Perform linear difference compensation: Processing module 30 obtains the current time. effective payload value and the predicted residual error obtained by performing the exponential decay operation. Based on the principle of linear superposition in signal processing, the effective payload value output by the sensor during the unloading phase... Essentially, it consists of two parts: one part is the actual physical load on the weighing pan (theoretically zero at this point), and the other part is the viscoelastic error component caused by the hysteresis of the elastomer material. Since the Kelvin-Voigt variable parameter model of this invention has independently predicted this viscoelastic error component, the processing module 30 executes the following difference synthesis formula: ; in, This represents the net weight output value after dynamic compensation. The value obtained by removing the system zero from the original sampled value, i.e., the effective payload value; This represents the residual error in the model's prediction. Through this step, processing module 30 separates the hysteretic viscoelastic error component from the total signal in the digital signal domain, thus... It can restore the actual force state on the weighing pan in real time, that is, achieve the technical effect that the physical state has not yet returned to zero, but the data shows that it has returned to zero.

[0056] Zero-point dead-zone clamping is performed. Although the above compensation algorithm can significantly reduce residual error, it is limited by the quantization noise of the analog-to-digital converter (ADC) and random environmental disturbances, resulting in a limited net weight output value. A slight positive or negative jump will occur near zero. Processing module 30 will output the net weight value. With the preset zero-point dead zone threshold Compare the zero-point dead zone threshold. The setting must comply with the requirements for zero-point error in metrological regulations (such as OIML R76 standard), and is usually set to the verification scale division value of the electronic scale. 0.25 to 0.5 times (i.e. ).like The processing module 30 determines that the current system is in an empty scale state and forces the output of the net weight value. The value is assigned to 0. This step uses hysteresis comparison logic to filter out invalid background noise, ensuring that the electronic scale displays a stable reading when unloaded, and avoiding frequent fluctuations in the last digit that could interfere with the user's reading.

[0057] Zero-point tracking and discrimination after compensation: Processing module 30 outputs the compensated net weight value. As input signal, the Automatic Zero Tracking (AZT) algorithm is run. Processing module 30 determines the net weight and outputs the value. Whether it remains within the zero-point tracking range for a continuous period of time (usually...) ,in (For displaying the scale division value). In traditional electronic scale logic, due to the slow natural physical zero-return process of the sensor (i.e., the long-tail effect), the original signal often exceeds the capture range of AZT for a long time, causing the system to fail to update the zero point in time, or misinterpreting slow creep recovery as a change in effective load. The innovation of this embodiment lies in: utilizing the net weight output value corrected by performing a linear differential compensation step. This discrimination mechanism allows the system to detect when a small amount of residual deformation still exists at the sensor physical level, indicating that the weighing system has entered a logically idle state. This mechanism ensures that the automatic zero-point tracking algorithm can intervene normally to correct the true reference drift caused by temperature changes or sensor temperature rise, achieving parallel and coordinated dynamic compensation for creep and zero-point tracking for temperature drift in the time domain.

[0058] Data normalization and output: Processing module 30 outputs the net weight value after clamping. The electronic scale is formatted according to its set range and graduation values. The processing module 30 maps the high-precision floating-point calculation results into discrete integer display values ​​based on the preset measurement mode (such as rounding mode or tail-cutting mode).

[0059] Finally, the processing module 30 transmits the normalized data to the display screen driver unit for numerical display, and sends the final weighing data to the external host computer through the communication interface, completing a complete weighing and compensation process.

[0060] See attached document Figure 2 To ensure that the variable-parameter Kelvin-Voigt model accurately reflects the physical characteristics of a specific sensor, it is necessary to obtain the model's fundamental elastic modulus through a standardized experimental procedure. and basic viscosity coefficient This process is typically performed during the final commissioning or periodic maintenance phase of the electronic scale production line. The specific calibration sub-steps are as follows: This step is based on the time-stress equivalence principle of linear viscoelasticity theory for polymer materials. Under extremely short pulsed loads, the polymer chains within the material do not have time to undergo complex conformational rearrangements, and their stress history integral is approximately zero. In this case, the mechanical behavior of the sensor can be approximated as a linear second-order system with constant parameters. Therefore, by utilizing the free decay response under short-time step excitation, the fundamental linear parameters can be decoupled from the complex nonlinear long-term rheological behavior, serving as a benchmark for subsequent dynamic correction.

[0061] Establish a constant test environment and initialize: Place the sensor or scale to be calibrated in a constant temperature test environment with temperature fluctuations of less than ±1℃ for at least 30 minutes to eliminate structural internal stress caused by thermal expansion and contraction. Processing module 30 enters engineering calibration mode. At this time, the system will bypass the dynamic compensation algorithm that executes the load monitoring and stress history integration steps to execute the residual error prediction and compensation output, and temporarily disable the automatic zero-point tracking (AZT) function to ensure that the acquired signal is the raw analog-to-digital conversion (ADC) data without algorithm modification.

[0062] Apply standard step excitation: Apply a standard weight load to the sensor weighing pan. To maximize the elastic rebound response of the material and suppress viscous flow accumulation, the loading process needs to simulate an ideal step signal, and the load holding time... The time should be strictly controlled, with the value range set between 5 and 15 seconds. Standard weight load. The recommended size is 50% to 80% of the full capacity of the electronic scale. This is to ensure the load holding time is reached. After a certain time, quickly remove the weights to allow the sensor to enter a free unloading recovery state.

[0063] Acquisition and preprocessing of free decay response data: Processing module 30 records the raw homing data after the unloading action is completed at a high sampling frequency allowed by the system (e.g., 100Hz). Recording time window length. This should cover the time required for the residual signal to decay to the noise floor (typically approximately 30 to 60 seconds). Processing module 30 acquires the raw sampling sequence within this time period. And read the empty scale reference value before loading. Extracting the pure residual response signal through subtraction. : ; The purpose of this step is to eliminate the effects of ADC zero-point offset and mechanical tare weight on the vertical displacement of the exponential decay curve, ensuring... It only characterizes the dynamic error component caused by viscoelastic hysteresis.

[0064] Perform log-linear regression analysis: Processing module 30 or the external calibration host computer analyzes the residual response information. Feature extraction was performed on the data sequence. Based on the fundamental equations of the Kelvin-Voigt model, under conditions of no stress history accumulation, the residual response after unloading follows a single exponential decay law: ; in, This represents the initial residual amplitude at the moment of unloading; Let be the fundamental time constant to be determined.

[0065] To solve using the linear least squares method Taking the natural logarithm of both sides of the above equation: ; To ensure fitting accuracy, processing module 30 only selects residual response information. Greater than 3 times the display scale value ( The effective data segment of the signal is used in the calculation to eliminate interference from quantization noise at the end of the signal.

[0066] Processing module 30 by time As the independent variable, Using as the dependent variable, a linear regression fit is performed to obtain the slope of the line, which is the slope of the line. Basic time constant That is, by the slope, that is, by the slope of the straight line. The negative value of the reciprocal is determined as follows: ; Calculate and store fundamental physical parameters: in obtaining the fundamental time constant Then, the processing module 30 determines the basic elastic modulus. and basic viscosity coefficient Considering that the elastic modulus only serves as a scaling factor in digital signal processing models, this embodiment employs a normalization method to normalize the basic elastic modulus. Set to a dimensionless constant of 1. Under this definition, the basic viscosity coefficient is... The value is directly determined by the following formula: ; Here Equivalent to a unit of time in physical dimensions, it represents the inherent relaxation rate of the sensor under unfatigue loading conditions.

[0067] Next, the processing module 30 uses the intercept term obtained from the log-linear regression analysis. (Right now (Logarithm of residual amplitude at time) to inversely solve for the initial residual amplitude. The residual error prediction formula based on this invention Under short-time rectangular pulse excitation in the calibration experiment, the stress history integral value is approximately: Therefore, the processing module 30 calculates and solidifies the ratio correction coefficient using the following formula. : ; in, The standard load value applied, This refers to the actual loading and holding time. Curing ratio correction factor. Used to calibrate the amplitude ratio between the theoretical model and the actual sensor output, ensuring that the predicted weight error amplitude during subsequent dynamic compensation is consistent with the actual physical quantity. Processing module 30 will calculate the results (Right now The system baseline parameters are written to the non-volatile parameter area of ​​storage module 40, and used to calculate dynamic parameters in subsequent unloading determination and model parameter correction steps. The initial cardinality.

[0068] See attached document Figure 2 After obtaining the basic time constant (Right now Subsequently, in order to quantify the degree to which the viscous characteristics of the sensor material drift with external energy input, it is necessary to determine the stress sensitivity factor through multiple sets of loading experiments with varying durations. .

[0069] This step is based on a simplified engineering model of nonlinear viscoelasticity. From a microscopic physical perspective, as the external load duration increases, the free volume within the polymer elastomer material redistributes, leading to a gradual increase in the slip resistance of the molecular chain segments (i.e., macroscopic viscosity). In this embodiment, this monotonically increasing hardening effect over time is linearly mapped to the drift rate of the time constant relative to the stress impulse. The processing module 30 or an external calibration device performs the following specific calibration sub-steps: Constructing a variable-duration gradient loading test sequence: Under the condition of keeping the ambient temperature constant, perform tests on the same sensor. This involves a series of independent load / unload loop tests. To ensure the validity of the statistical patterns, The value ranges from 3 to 5. In each test... (in In the same standard weight load, apply the same standard weight load. However, different load retention times can be set. Load hold time The distribution should follow a logarithmic or geometric progression (e.g., set to 60s, 300s, 900s, 1800s respectively) to cover the typical operating range of electronic scales from short-term weighing to long-term monitoring.

[0070] To eliminate the interference of elastic aftereffects on calibration data, the system must execute a strict zeroing wait procedure between every two adjacent tests. The processing module 30 monitors the sensor output, and only when the output value stabilizes at the initial zero point for a continuous period (e.g., 30 seconds) will the system wait for the value to return to zero. of Within the specified range, indicating that the internal microstructure of the material has returned to thermodynamic equilibrium, the next round of testing can be initiated.

[0071] Obtain the equivalent time constant for each test group: for each test The processing module 30 calls the discrete integral algorithm to calculate the historical stress integral value during the loading process in real time. When the specified holding time is reached... At the instant the unloading action is executed, the processing module 30 locks the accumulated points value at this moment and records it as follows. Simultaneously, the processing module 30 collects the residual decay curve after unloading and uses the log-linear regression method to calculate the observation time constant under this long-term loading condition. The observation time constant measured at this time Characterizes the sensor after experiencing The instantaneous rheological state after a stress history of magnitude.

[0072] Constructing a parameter-differentiated dataset: Processing module 30 processes the cumulative integral values ​​obtained from each group of tests. As an independent variable The relative rate of change of the observed time constant with respect to the baseline time constant is used as the dependent variable. Dependent variable The calculation formula is as follows: ; This formula eliminates the differences in the basic stiffness of individual sensors, retaining only the proportional characteristics of viscous drift. The processing module 30 then generates a set of feature point pairs for fitting. .

[0073] The least squares fitting method passing through the origin is performed to solve the problem: According to the physical model established in this invention, when the stress history integral value is zero, the sensor's time constant should strictly regress to the fundamental time constant. That is, there is no intercept term. Therefore, processing module 30 adopts a linear regression model that is forced to pass through the origin. The solution is performed, and the optimal fitting slope is the stress sensitivity factor to be determined. The calculation formula is as follows: ; in, As a stress-sensitive factor, if The unit is kg·s, then The unit is 1 / (kg·s). For commonly used aluminum alloy elastomer sensors, The typical value range is 1.0 × 10. 5 Up to 5.0×10 4 .

[0074] Parameter verification and consolidation: Processing module 30 calculates the coefficient of determination for regression analysis. .like This indicates that the rheological properties of the sensor highly conform to the linear variable parameter assumption of the present invention. The processing module 30 will calculate the... Write it into non-volatile memory as the actual operating parameters. If This indicates that the sensor may have material defects or have undergone nonlinear plastic deformation. The processing module 30 outputs a calibration failure alarm and disables the dynamic compensation function to prevent model mismatch from amplifying the weighing error.

[0075] Specific application examples: To further verify the effectiveness of this invention in solving the pain point of rapid, small-amount weighing of heavy items after prolonged storage in real-world commercial scenarios, this embodiment constructs a typical application scenario in a supermarket fresh produce section, and combines... Figures 3-5 The simulated waveforms shown are explained in detail.

[0076] Scene presets and basic parameter configurations: This embodiment simulates a high-temperature summer environment (higher ambient temperatures intensify the movement of polymer chain segments in elastomer materials, leading to enhanced viscous characteristics). The parameters of the intelligent electronic scale used are as follows: Maximum range Scale value of verification The basic physical parameters obtained by the system beforehand through the calibration procedure are as follows: Basic relaxation time constant ; Stress-sensitive factors ; Residual amplitude coefficient .

[0077] Phase 1: Large loads and long-term loading (stress history memory process) such as Figure 3 The timing diagram shown is on the time axis. to During the period, a large load event occurred. Specifically, a customer placed a 10kg watermelon on the scale pan to select other items, and the load was maintained for 5 minutes (300s).

[0078] During this period, the processing module continuously monitors that the effective payload value is greater than the loading threshold, and the integration register remains active. The system performs trapezoidal integration. Figure 3 The area of ​​the gray-filled region represents the numerical accumulation process within the integration register.

[0079] when The moment just before the unloading action occurs, the final integral value recorded in the integral register. The calculation is as follows: ; This value quantifies the deep creep history that occurred inside the sensor's elastomer. Physically, this means that the material's molecular chain segments have undergone significant slippage and orientation, generating substantial mechanical hysteresis potential energy.

[0080] Phase Two: Unloading and Dynamic Solution of Model Parameters (Parameter Correction Process): When a customer removes a watermelon, the processing module detects a significant negative slope in the original sampled value, determines that an unloading event has occurred, and immediately locks the final integral value. .

[0081] like Figure 4 As shown in the bar chart, the system uses the variable parameter model of this invention to dynamically correct the relaxation time constant. The base relaxation time constant before correction is... The time was 4.0 seconds (blue column). The system operates according to the formula... Calculate the dynamic relaxation time constant in the current state: ; Corrected dynamic relaxation time constant It becomes 4.6s (orange bar). The +15% marked in the figure intuitively reflects that due to prolonged heavy pressure, the sensor material exhibits physical properties of hardening or becoming stickier, resulting in its zero-return speed being 15% slower than the standard state.

[0082] At the same time, the system calculates the instant of unloading ( The initial residual error quantity of the theory exists : ; This means that although the weighing pan is empty, the sensor still physically retains a false reading of about 20g.

[0083] Phase Three: Rapid continuous weighing (superimposed compensation process) such as Figure 5 The signal comparison chart shows the time required after unloading within a very short period of time. to The system response is as follows: The specific action is: only 1.0 second after the customer removes the watermelon (i.e., ... At that moment, the salesperson quickly placed a box of grapes with an actual weight of 200g (0.200kg) onto the scale. The graph shows a comparison of three key curves: The red dashed line (raw sensor signal) represents the uncompensated signal acquired by the analog-to-digital converter (ADC). Due to the hysteresis caused by the watermelon not yet dissipating, in... At that moment, the original signal is essentially the superposition of the grape weight and the residual error of the watermelon. According to the law of exponential decay, the residual error of the watermelon at this point has only decreased to approximately 16g (20). Therefore, the ADC reading is as high as If it were a traditional electronic scale, the display would show 0.215kg or 0.220kg, directly leading to an overcharge of 15g to 20g due to measurement inaccuracy.

[0084] The solid blue line (system output net weight) represents the output result after compensation by the processing module of this invention. The system is based on the elapsed time. and dynamic parameters Real-time calculation of prediction residual error : ; The processing module performs the difference operation: As shown in the figure, the blue solid line is... It remained stable at the 200g mark, eliminating residual errors in the substrate.

[0085] Black dashed line (actual weight): Represents the true physical value and is used for reference.

[0086] Implementation Results Summary: Through Figure 5 The comparison in the lower right corner shows that although the physical sensor did not achieve zeroing (the red dotted line is still above zero), at the user-perceived display level (blue solid line), the system successfully eliminated the lag effect caused by the previous large order (watermelon). This invention enables the electronic scale to maintain the measurement accuracy at the calibration graduation level in scenarios of rapid, small-amount weighing of heavy objects left for extended periods, avoiding disputes over short weights in commercial trade.

Claims

1. A parameter management system for intelligent commercial electronic scales with an integrated data processing module, characterized in that, include: A conversion module (20) is connected to a sensor (10) for converting physical pressure into an analog voltage signal and outputting digital raw sampled values; Storage module (40) stores the basic elastic modulus, basic viscosity coefficient and stress sensitivity factor; Processing module (30), the processing module (30) is configured to: When the sensor (10) is detected to be in a loaded state, the product of the load amplitude determined based on the original sampled value and the sampling time interval is calculated, and the calculation result is updated to the integration register (31) to record the stress history integration value; When an unloading event is detected in the sensor (10), the basic viscosity coefficient is corrected according to the linear function relationship between the stress history integral value and the stress sensitivity factor, and the dynamic viscosity coefficient corresponding to the current weighing is calculated. The dynamic relaxation time constant is determined by the ratio of the dynamic viscosity coefficient to the basic elastic modulus, and the predicted residual error, which decays exponentially with time, is calculated based on the dynamic relaxation time constant. The corrected net weight value is obtained by subtracting the predicted residual error from the original sampled value and outputting it as the final weighing result.

2. The intelligent commercial electronic scale parameter management system with integrated data processing module according to claim 1, characterized in that, The calculation is based on the product of the load amplitude determined by the original sampled value and the sampling time interval, and the calculation result is updated to the integration register (31) to record the stress history integration value. Specifically, this includes: Read the system zero point value maintained in the storage module (40), subtract the system zero point value from the original sampled value to obtain the effective load value, and use the effective load value as the load amplitude; The effective payload value is compared with a preset loading threshold. If the effective payload value is continuously greater than the loading threshold, the sensor (10) is determined to be in the loading state. The trapezoidal numerical integration algorithm is used to calculate the single-step stress increment using the effective load value at the current sampling time, the effective load value at the previous sampling time, and the actual sampling time interval, and the single-step stress increment is accumulated to the integration register (31).

3. The intelligent commercial electronic scale parameter management system with integrated data processing module according to claim 1, characterized in that, The correction of the basic viscosity coefficient based on the linear functional relationship between the stress history integral value and the stress sensitivity factor specifically includes: Calculate the product of the stress history integral value and the stress sensitivity factor, and add the product to a constant 1 to obtain the dynamic correction ratio; The dynamic viscosity coefficient is obtained by multiplying the basic viscosity coefficient by the dynamic correction ratio, wherein the dynamic viscosity coefficient is used to characterize the macroscopic viscous resistance of the sensor material after experiencing stress history.

4. The intelligent commercial electronic scale parameter management system with integrated data processing module according to claim 1, characterized in that, The calculation of the exponentially decaying prediction residual error based on the dynamic relaxation time constant specifically includes: Record the unloading moment when the unloading event occurs, and use the stress history integral value and the dynamic relaxation time constant to calculate the initial residual amplitude at the moment of unloading; Obtain the absolute elapsed time of the current moment relative to the unloading moment, and determine the prediction residual error at the current moment according to the natural exponential decay law based on the initial residual amplitude, the absolute elapsed time, and the dynamic relaxation time constant. When the processing module (30) detects the loading state again, it clears the value of the integration register (31).

5. The intelligent commercial electronic scale parameter management system with integrated data processing module according to claim 2, characterized in that, The processing module (30) is also configured to perform integral saturation protection: A preset upper limit for integration is set, which is determined based on the product of the maximum range of the electronic scale and the time required for the sensor material to reach creep saturation. After each update of the integration register (31), the processing module (30) determines whether the stress history integration value exceeds the integration upper limit value; If the stress history integral value exceeds the integral upper limit value, the value of the integral register (31) will be forcibly set to the integral upper limit value.

6. The intelligent commercial electronic scale parameter management system with integrated data processing module according to claim 1, characterized in that, The processing module (30) monitors the uninstallation event in the following way: Calculate the rate of change of the original sampled value within a preset sliding window, and compare the rate of change with a preset negative unloading slope threshold; When the rate of change is less than the unloading slope threshold and the duration exceeds the anti-jitter cycle, it is determined that the sensor (10) has experienced the unloading event, and write operations to the integration register (31) are prohibited.

7. The intelligent commercial electronic scale parameter management system with integrated data processing module according to claim 1, characterized in that, After obtaining the corrected net weight value and outputting it as the final weighing result, the following is also included: The net weight value is compared with a preset zero-point dead zone threshold. If the net weight value is less than or equal to the zero dead zone threshold, then the net weight value is forcibly assigned to zero; The net weight value after forced assignment is used as the input signal to run an automatic zero-point tracking algorithm to correct the zero-point drift of the system when there is residual deformation at the physical level of the sensor (10).

8. The intelligent commercial electronic scale parameter management system with integrated data processing module according to claim 1, characterized in that, The basic elastic modulus and the basic viscosity coefficient are obtained in the following manner: A short-time standard step excitation is applied to the sensor (10) under constant temperature conditions, and the load magnitude of the standard step excitation is recorded. At the instant of applying the standard step excitation, the instantaneous response amplitude of the sensor (10) is acquired, and the ratio of the load magnitude to the instantaneous response amplitude is calculated as the basic elastic modulus. After removing the short-time standard step excitation, free decay response data is collected, residual response signal is extracted from the free decay response data, and log-linear regression analysis is performed on the relationship between the residual response signal and time. The basic time constant is determined based on the slope of the regression line, and the basic time constant is then used as the basic viscosity coefficient.

9. A parameter management system for an intelligent commercial electronic scale with an integrated data processing module as described in claim 8, characterized in that, The stress sensitivity factor is obtained in the following manner: Multiple cyclic tests were performed on the sensor (10), with the same standard load applied in each cyclic test but different load holding times set to form a variable duration gradient loading test sequence; For each loop test, record the cumulative integral value at the moment of unloading, and calculate the observation time constant after unloading; Construct a dataset in which the cumulative integral value is the independent variable and the rate of change of the observed time constant relative to the basic time constant is the dependent variable; The dataset is fitted using the least squares method passing through the origin, and the slope obtained from the fitting is determined as the stress sensitivity factor.

10. A parameter management system for an intelligent commercial electronic scale with an integrated data processing module according to claim 4, characterized in that, The determination of the prediction residual error at the current moment according to the natural exponential decay law specifically includes: The single-step attenuation factor is pre-calculated based on the dynamic relaxation time constant and the system sampling period; The predicted residual error calculated at the previous sampling time is multiplied by the single-step decay factor to obtain the predicted residual error at the current time.

Images