Data acquisition method and system based on digital metering technology

Abstract

The invention relates to the technical field of industrial automation control, in particular to a data acquisition method and system based on a digital metering technology, and the method sequentially comprises the following steps: 1, carrying out the dynamic calibration of a sensor; 2, constructing a phase compensation clock tree; 3, establishing a temperature-humidity-vibration three-factor coupled error compensation model, determining the interaction weight of each factor through an orthogonal test method, and updating a compensation coefficient on line by adopting a recursive least square method with a forgetting factor; and step 4, generating a transmission priority according to the product of the data change rate and the health degree index, dynamically allocating a transmission bandwidth based on a network quality evaluation value calculated based on the time delay change rate and the packet loss rate, and performing compressed coding of a lead code mark on high-priority data. The method not only improves the precision of the system, but also optimizes the data processing and transmission efficiency, and provides an effective solution for high-precision data acquisition in a large-scale and complex environment.

Description

Technical Field

[0001] This invention relates to the technical field of industrial automation control, and in particular to a data acquisition method and system based on digital metering technology. Background Technology

[0002] In the field of industrial automation control, especially in high-precision manufacturing and precision testing, multi-sensor data acquisition technology is the core foundation for real-time monitoring and process control. Existing technologies have revealed the following key shortcomings in practical applications:

[0003] Existing synchronous acquisition solutions mostly rely on hardware clock distribution circuits, using a fixed-frequency clock signal to trigger sampling from multiple sensors. These solutions do not consider signal delay differences caused by cable impedance mismatches during long-distance transmission, nor do they address clock jitter issues caused by electromagnetic interference. For example, in an automotive welding production line, the time stamp deviation between force sensors and vision sensors at different workstations can reach millisecond levels due to clock asynchrony. This results in a spatial matching error exceeding ±0.5mm between the robotic arm's movement trajectory and the welding point position, directly impacting welding quality.

[0004] Traditional environmental compensation methods typically design linear compensation models for single environmental factors (such as temperature or vibration), neglecting multi-physics coupling effects. Taking injection molding machine pressure sensors as an example, existing solutions only compensate for mold temperature, but fail to consider the nonlinear deformation effect of hydraulic oil temperature fluctuations on the sensor's elastomer, while also ignoring the signal modulation effect caused by high-frequency vibration. Experiments show that when the temperature changes by 10°C and the vibration frequency exceeds 200Hz, the residual error of traditional compensation methods can reach 2.8 times the nominal accuracy, resulting in excessive thickness uniformity in the finished product.

[0005] In summary, existing technical solutions can no longer meet the stringent requirements of intelligent manufacturing for data acquisition systems, and there is an urgent need for an innovative technical solution that can systematically solve the above problems. Summary of the Invention

[0006] To achieve the above objectives, the present invention provides a data acquisition method and system based on digital metering technology, wherein the data acquisition method based on digital metering technology performs the following steps in sequence:

[0007] Step 1: Perform dynamic calibration of the sensor. By continuously collecting temperature drift curves, response time change rate and zero offset from the sensor's historical working data, calculate the coefficient of variation of each parameter to generate a real-time health index. Based on the correlation between this real-time health index and the real-time detected ambient temperature and humidity, dynamically adjust the sensor's reference parameters.

[0008] Step 2: Construct a phase-compensated clock tree, generate branch clock signals with forward prediction and feedback correction mechanisms based on the main controller clock, and collect physical quantity signals and corresponding sensor housing temperature gradient, power supply voltage fluctuation extreme environmental parameters under synchronous clock control.

[0009] Step 3: Establish an error compensation model that couples temperature, humidity and vibration, determine the interaction weights of each factor through orthogonal experiment, and update the compensation coefficients online using the recursive least squares method with a forgetting factor.

[0010] Step 4: Generate transmission priority based on the product of data change rate and health index, dynamically allocate transmission bandwidth based on network quality assessment value calculated based on latency change rate and packet loss rate, and implement preamble-marked compression encoding on high priority data.

[0011] Accordingly, embodiments of the present invention also provide a data acquisition system based on digital metering technology, used to run a data acquisition method based on digital metering technology as described in any embodiment of the present invention, including:

[0012] The sensor condition assessment module includes:

[0013] The temperature drift analysis unit is electrically connected to the temperature probe on the sensor housing and is used to collect and store the temperature-zero offset correspondence.

[0014] The response time monitoring unit has a standard signal generator connected to its input and a timer connected to its output to calculate the rate of change of response time.

[0015] The health calculation unit has its input terminals connected to the temperature drift analysis unit, response time monitoring unit, and zero-point calibration unit, respectively, and its output terminal connected to the parameter adjustment module via a data bus.

[0016] The clock synchronization module includes:

[0017] The phase prediction subunit is directly connected to the main controller clock source and has a built-in autoregressive model calculation unit.

[0018] The feedback correction subunit has at least three historical clock offset registers connected to its input and a clock divider connected to its output via a digital-to-analog converter.

[0019] A branch clock generator, the output of which is connected to the trigger interfaces of each sensor through an isolation circuit;

[0020] The environmental coupling compensation module includes:

[0021] The multi-parameter acquisition unit includes a temperature sensor array, a vibration accelerometer, and a voltage monitoring circuit, and is connected to the main controller via a parallel data interface.

[0022] The weight calculation engine connects to the orthogonal experimental database and real-time environmental parameter cache at the input end and to the compensation matrix update unit at the output end.

[0023] An online compensation actuator has its data input end connected to the sensor signal channel and its control end connected to the compensation matrix update unit.

[0024] The data transmission control module includes:

[0025] The priority calculation unit has inputs connected to the data change rate analyzer and the health assessment unit, and integrates a dynamic weight allocator.

[0026] The bandwidth adaptive unit includes a network quality monitoring circuit and a compression code selector. The network quality monitoring circuit obtains the link status in real time through the physical layer interface.

[0027] The redundant control unit has at least three independent network interfaces connected to its output and a priority tag generator connected to its input.

[0028] The beneficial effects of this invention are:

[0029] Dynamic sensor calibration ensures measurement accuracy, a phase-compensated clock tree guarantees data acquisition synchronization, and an error compensation model dynamically corrects for the impact of environmental changes on the data. Finally, a data transmission priority and dynamic bandwidth allocation strategy ensures efficient use of network resources, avoiding packet loss or latency issues when dealing with large volumes of data. Overall, this approach not only improves system accuracy but also optimizes data processing and transmission efficiency, providing an effective solution for high-precision data acquisition in large-scale, complex environments. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 This is a flowchart of the steps of the method of the present invention;

[0032] Figure 2 This is a flowchart of the orthogonal experimental method in step 3 of the method of the present invention;

[0033] Figure 3 This is a structural block diagram of the system of the present invention. Detailed Implementation

[0034] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should also be noted that, to make the embodiments more comprehensive, the following embodiments are the best and preferred embodiments, and those skilled in the art can use other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0035] Please see Figures 1-3 This invention provides a data acquisition method based on digital metrology technology, wherein step 1 involves performing dynamic calibration of the sensor. During this process, the sensor's temperature drift curve, response time change rate, and zero-point offset are continuously collected, and a real-time health index is generated by calculating the coefficient of variation of each parameter. The health index is a quantitative assessment of the sensor's state; it combines real-time detected ambient temperature and humidity data, and dynamically adjusts the sensor's baseline parameters based on their correlation with the health index. This step can correct sensor deviations in real time, ensuring the sensor's accuracy under various environmental conditions and reducing measurement errors caused by changes in environmental factors.

[0036] In step 2, a phase-compensated clock tree is constructed to ensure the synchronous acquisition of physical quantity signals and environmental parameter data. Using the main controller clock as a reference, branch clock signals with forward prediction and feedback correction mechanisms are generated to precisely synchronize the sampling times of each data channel. Clock synchronization is a critical step in data acquisition, especially in multi-sensor systems where temporal consistency is paramount. This clock synchronization technique eliminates data timing problems caused by clock skew, improving the accuracy of the acquired data.

[0037] Step 3 involves establishing an error compensation model that couples temperature, humidity, and vibration. The interaction weights of each factor are determined using orthogonal experimental design, and the compensation coefficients are updated online using a recursive least squares method with a forgetting factor. This step allows for real-time dynamic updates to the error compensation model based on environmental changes, ensuring high accuracy of sensor measurements even under the influence of multiple environmental factors such as temperature, humidity, and vibration. This error compensation mechanism plays a crucial role in improving the long-term stability of the system and the accuracy of the data.

[0038] The final step is to generate data transmission priorities based on the product of the data change rate and the health index, and dynamically allocate transmission bandwidth based on network quality assessment values ​​(including latency change rate and packet loss rate). High-priority data will be compressed and encoded with preamble tags to optimize network resource utilization and ensure the timely transmission of important data. This strategy effectively addresses network congestion and ensures the stability and real-time performance of the system in a multi-tasking environment.

[0039] In one possible implementation, a reference curve is generated by recording the zero-point offset at preset temperature intervals in a constant-temperature environment. The key to this step is constructing the sensor's behavior pattern at different temperatures by establishing the correspondence between the sensor housing temperature and the zero-point offset. The reference curve is obtained by measuring the zero-point offset at a preset constant temperature. These data serve as a reference in practical applications. Through interpolation, the expected offset at the current temperature can be obtained from the reference curve using the housing temperature detected by the temperature sensor during real-time operation. This step ensures accurate prediction and compensation of the zero-point offset when the temperature changes, thereby improving the sensor's measurement accuracy under different temperature conditions.

[0040] The rate of change of response time is a crucial parameter of a sensor's response characteristics, describing how quickly the sensor reacts to changes in the input signal. In practice, the sensor is first excited by a standard step signal. Then, the time difference required for the response time to reach 10% to 90% of its steady-state value is measured, and this time difference is calculated as the rate of change of response time. This process is updated once per sampling period to track changes in the sensor's response speed in real time. By periodically updating the rate of change of response time, fluctuations in the sensor's response capability can be effectively monitored and compensated for, ensuring a stable and reliable response speed under different operating conditions.

[0041] Zero-point offset measurement is performed when there is no input signal, typically by acquiring the sensor's output value in a no-input state. To eliminate the influence of random noise, a sliding window averaging method can be used to smooth the acquired data. This method removes accidental noise by calculating the average value of the data within a certain time window, obtaining a more stable and accurate zero-point offset. This step effectively eliminates the influence of external interference on the sensor output, ensuring stable zero-point data for each measurement, thereby reducing errors.

[0042] In one possible implementation, firstly, a two-dimensional lookup table of compensation coefficients, incorporating temperature and humidity variations, needs to be established. This lookup table is derived from extensive experimental data or model derivation, recording the compensation coefficients under different temperature and humidity combinations. This lookup table can reflect the impact of temperature and humidity changes on sensor performance in real time, providing basic compensation information.

[0043] In practical applications, the temperature and humidity of the sensor's environment are constantly changing, necessitating the use of a lookup table to obtain the corresponding compensation amount. Bilinear interpolation allows the system to find the appropriate compensation coefficient from the lookup table based on the current temperature and humidity values. By interpolating the coordinates of temperature and humidity, bilinear interpolation accurately obtains the baseline compensation amount under the current environmental conditions. This step ensures that the system can accurately calculate suitable compensation coefficients under different environmental conditions, thereby better calibrating the sensor's measurements.

[0044] The health index, reflecting the sensor's condition, typically ranges from 0 to 1. To use the health index for calibration, it is first converted into a reliability factor, which reflects the sensor's health status and measurement reliability. The reliability factor is calculated using a piecewise linear transformation based on a preset interval of the health index. This allows for a larger calibration impact when the health index is high, while reducing its effect when it is low, avoiding over-calibration due to sensor malfunction or performance degradation.

[0045] By combining the aforementioned confidence factor with the base compensation amount, the calibrated parameters are finally calculated using a formula. The specific calculation method is: Calibration parameter = Original parameter × (1 + Confidence factor × Base compensation amount). This method combines real-time environmental changes and sensor health status, enabling dynamic adjustment of the baseline parameters. When the health index is high, the confidence factor is large, and correspondingly, the base compensation amount has a greater impact on the parameters, resulting in more accurate compensation. Conversely, when the health index is low, the confidence factor is small, and the impact of the compensation amount on the parameters is reduced, thus avoiding inaccurate compensation.

[0046] In one possible implementation, the clock tree is first constructed based on the clock frequency of the pre-controller as a reference. This clock frequency is typically provided by the system master clock and serves as the reference for the entire system. An initial division factor is determined based on the maximum sampling rate of the sensor required by the system. The division factor determines the frequency of the output clock; generally, to ensure efficient sensor operation, the division factor should be precisely set according to the sensor's maximum sampling rate to guarantee the stability and efficiency of the clock signal.

[0047] In clock tree design, predicting clock offset is a crucial step. Offset refers to the difference between the actual clock signal and the desired clock signal over consecutive clock cycles. To address this, an autoregressive model can be used to predict the clock offset for the next cycle. The order of the autoregressive model is determined by the number of zero-crossing points in the autocorrelation function of historical offset data. By using past offset data for prediction, the autoregressive model can more accurately estimate future clock offsets, allowing for timely adjustments in advance. This method effectively reduces clock fluctuations and ensures clock signal stability.

[0048] In practical applications, clock signals may be affected by various external factors, leading to periodic offsets. Therefore, a feedback correction stage is necessary. By collecting the actual offsets of the most recent three clock cycles, the current clock error can be assessed. A weighted average offset is calculated by assigning decreasing weights based on the time elapsed. Weighted averaging effectively reduces the impact of short-term noise on the correction results, focusing more on recent offset changes and making the correction process more accurate.

[0049] After feedback correction, the next step is to dynamically adjust the frequency division factor so that the clock jitter value gradually decreases exponentially until the clock offset stabilizes within a preset allowable range. This process reduces clock jitter by continuously optimizing the frequency division factor, ensuring that the clock frequency tends to stabilize and meets the system's accuracy requirements.

[0050] In one possible implementation, to ensure strict synchronization between environmental parameter acquisition and physical quantity sampling, the rising edge of the branch clock is used as a trigger signal to activate both the physical quantity sampling channel and the environmental parameter acquisition channel. This means that each rising edge of the branch clock simultaneously controls the activation of multiple channels, ensuring that the acquisition of physical quantities (such as temperature, vibration, etc.) and environmental parameters (such as voltage, external temperature, etc.) occurs at the same time, thereby avoiding data inconsistencies caused by timing errors. Synchronous sampling helps improve measurement accuracy, especially when multiple environmental factors may simultaneously affect physical quantities.

[0051] Voltage fluctuations are a significant factor affecting the stability of digital systems, making the acquisition of extreme values ​​of supply voltage fluctuations crucial. In this process, a peak hold circuit is employed to capture the maximum and minimum voltage values. The peak hold circuit records the instantaneous peak voltage in real time and holds it for a period of time for subsequent analysis. This design effectively avoids the interference of instantaneous voltage fluctuations on system performance, providing accurate data for subsequent power supply stability assessments.

[0052] Variations in the casing temperature gradient have a significant impact on equipment performance, especially in high-precision sensor applications, where temperature fluctuations can lead to measurement errors. Therefore, by arranging three temperature sensors along the sensor axis, temperature measurements are acquired at different locations. These data can be used to calculate the linear rate of temperature change, thereby deriving the casing temperature gradient. This temperature measurement method, through the coordinated operation of multiple sensors, can accurately capture changes in the external ambient temperature, thus providing reliable temperature compensation for the measurement results of physical quantities.

[0053] Vibration is a common environmental factor affecting equipment operation, making the extraction of vibration spectral characteristics crucial. By acquiring vibration signals using accelerometers and performing spectral analysis using Fast Fourier Transform (FFT), the dominant frequency components of the vibration signals can be extracted. These dominant frequency components reflect the vibration characteristics of the equipment, such as whether abnormal vibrations exist or whether it is in a resonant state within a specific frequency range. This method provides an accurate vibration environment background for physical quantity acquisition and further analyzes the impact of vibration on sensor or equipment performance.

[0054] By maintaining the peak value of voltage fluctuations, accurately calculating temperature gradients, and extracting vibration spectrum characteristics, the entire acquisition process not only reflects the equipment's operating status in real time but also effectively compensates for measurement results under environmental changes (such as voltage fluctuations, temperature changes, and vibration). Thus, through synchronous sampling and detailed recording of environmental factors, the system's anti-interference capability and data accuracy are improved, ultimately providing reliable support for high-precision digital metering, ensuring stable system operation under various environmental conditions, and reducing errors and risks.

[0055] In one possible implementation, in step 3, orthogonal experimental design is used to analyze the impact of environmental factors on sensor output deviation. Specifically, a full-factor experiment is conducted by setting three environmental factors—temperature, humidity, and vibration—at low, medium, and high test levels, respectively. Each test level represents an environmental condition, and different combinations of temperature, humidity, and vibration can simulate various environmental situations that may be encountered in practical applications. Through experiments, the sensor output deviation is measured under each environmental combination, accumulating data on the impact of each factor and its combinations.

[0056] After collecting experimental data, analysis of variance (ANOVA) was used to separate the main effects (i.e., individual effects) of each factor and their interaction effects (i.e., the influence of interactions between two or more factors). ANOVA calculates the contribution of different environmental factors to the output deviation, clarifying the degree of influence of each factor on the sensor deviation. By assigning weights to the main effects, it is possible to determine which factors have the greatest impact on the sensor output and which are less significant.

[0057] Based on analysis of variance, the weights of main effects are allocated proportionally to the contribution of each factor to the output bias. Meanwhile, the weights of interaction effects are determined by the proportion of the square of the effect value to the total sum of squares. In this way, by allocating weights to main and interaction effects, the influence of each factor on sensor bias under different environmental conditions can be scientifically assessed, thereby optimizing sensor calibration and compensation methods.

[0058] To further improve the system's adaptability, a forgetting factor is introduced into the data acquisition method. The forgetting factor is dynamically adjusted based on the rate of change of environmental parameters. When the rate of change of environmental parameters exceeds a preset threshold, the system automatically increases the value of the forgetting factor, ignoring the influence of past data to respond more quickly to new environmental changes. This method adapts to environmental changes by rapidly adjusting the forgetting factor, thereby ensuring the real-time nature and accuracy of data acquisition. Dynamic adjustment of the forgetting factor ensures the system's sensitivity to sudden environmental changes and reduces data obsolescence or bias caused by long-term stable environmental changes.

[0059] In one possible implementation, during the transmission priority generation process in step 4, the data change rate is first normalized. The data change rate refers to the difference between adjacent sampling points, which reflects the rate of change in the sensor output. To make different data change rates comparable, the moving average of the differences between adjacent sampling points is first calculated and then divided by the sensor's range. The range reflects the sensor's maximum measurable range. Through this normalization process, the output change rates of different sensors can be converted into a uniform scale, thereby effectively comparing the speed of data change.

[0060] The health index is an important indicator reflecting the health status of a sensor, and is usually related to factors such as sensor aging and damage. To make the health index more aligned with transmission priority, non-linear compression processing is employed. Specifically, a segmented processing method is used based on the numerical range of the health index:

[0061] When the health index falls below a first threshold, a logarithmic transformation is used. This is because when the health index is poor, the sensor's performance drops sharply, and using a logarithmic transformation can enhance the sensitivity to changes in sensor performance.

[0062] When the health index exceeds the second threshold, a square root transformation is used. This smooths out changes in the higher health index, preventing sensors with better health from being given excessively high priority due to slight fluctuations.

[0063] This non-linear compression can more accurately reflect the impact of the health index on priority generation, thereby avoiding sensors with poor health from being assigned too low a priority.

[0064] Ultimately, priority values ​​are generated by weighted summation of the normalized data change rate and the transformed health index. Dynamic weights are used to balance the contributions of both. These dynamic weights are adjusted linearly between 0.3 and 0.7 based on the current network congestion level. When network congestion is light, priority may depend more on the data change rate, while when network congestion is heavy, priority is more influenced by the health index, thus prioritizing the transmission of sensor data with better health status.

[0065] In one possible implementation, the latency variation rate is a metric for measuring network latency volatility. To accurately assess latency variation, test packets are first continuously sent, and the round-trip time (RTT) from transmission to reception is measured. The latency variation rate is obtained by calculating the standard deviation of these RTTs. The standard deviation reflects latency volatility; a higher standard deviation indicates significant latency fluctuations in the network, potentially affecting data transmission stability. In this way, the latency variation rate can serve as a key indicator for bandwidth allocation in subsequent decisions.

[0066] Packet loss rate directly impacts data transmission reliability. In this method, the packet loss rate is calculated using a variable-size window counting method. Specifically, the packet loss rate is calculated based on the number of packets lost within the window, and the window size is variable, with the window duration being an integer multiple of the current data sampling period. This makes the packet loss rate calculation more flexible and allows for adjustments to the evaluation accuracy based on different network conditions and data sampling periods. When the packet loss rate is high, the system adjusts its bandwidth allocation strategy accordingly to ensure data transmission reliability.

[0067] Based on the statistical results of latency change rate and packet loss rate, and combined with their joint evaluation value, a preset strategy table can be consulted to select a suitable compression algorithm. This step considers network latency and packet loss; the evaluation value reflects the current state of the network and provides a basis for bandwidth allocation. When the joint latency-packet loss rate evaluation value is in a critical region (i.e., the network load is high or instability occurs), the system will activate the hybrid compression mode.

[0068] When network conditions are poor, especially with high latency or packet loss, the system selects an appropriate compression algorithm based on a preset strategy to optimize bandwidth utilization. In hybrid compression mode, the system uses lossless compression for the data packet header and lossy compression for the data payload. Lossless compression ensures the integrity of the data header information, while lossy compression reduces the payload by losing some data, thereby saving bandwidth. This flexible compression strategy can reduce network load while ensuring data transmission efficiency, making it particularly suitable for environments with limited bandwidth.

[0069] In one possible implementation, the system first calculates a dynamic threshold using a health index. The formula for calculating the dynamic threshold is: Dynamic threshold = Base threshold × (1 + Health index decay coefficient), where the base threshold is a preset standard value, the health index reflects the health status of the sensor, and the health index decay coefficient is obtained from a preset parameter table based on the sensor type.

[0070] Basic threshold: This value represents the standard threshold within the normal fluctuation range of sensor data.

[0071] Health Index Decay Coefficient: This coefficient is adjusted based on the sensor's operating condition and environmental changes. For example, a sensor's performance may degrade due to environmental factors or usage time, thus affecting the accuracy of the data.

[0072] When the data jump variable between adjacent sampling points exceeds the calculated dynamic threshold, it indicates abnormal data fluctuations, possibly caused by equipment malfunction, environmental interference, or other reasons. In this case, the system will trigger a multi-stage data repair process. The data jump variable refers to the change range between two adjacent data points; if it exceeds the dynamic threshold, it indicates an anomaly.

[0073] The data recovery process consists of three main stages:

[0074] In the first stage, the system uses an autoregressive moving average (ARMA) model to predict the data. An ARMA model is a time series analysis method that can predict future data trends using historical data. Under normal circumstances, the ARMA model can accurately predict data trends, thereby identifying outliers in the data.

[0075] If the predicted residual (i.e. the deviation between the predicted value and the actual value) exceeds the allowable range, it indicates that there is an error in the prediction result, and the data may have been affected by abnormal factors.

[0076] When the ARMA model's predictions are inaccurate, the system switches to the second stage, using data from associated sensors for spatial interpolation compensation. Spatial interpolation utilizes data from other sensors related to the target sensor's location to infer missing or anomalous data values. For example, if a sensor at a certain location malfunctions, the system can use data from sensors at neighboring locations for interpolation to supplement the anomalous data.

[0077] By using correlated sensor data for compensation, abnormal data can be spatially repaired, ensuring data continuity and accuracy.

[0078] Finally, the system generates an anomaly report, which includes the original data, the repaired data, environmental parameters, and a health index. The anomaly report details the entire data repair process, including anomaly detection, remedial measures, and their effects. This report provides crucial information for subsequent analysis, monitoring, and system optimization.

[0079] Accordingly, embodiments of the present invention also provide a data acquisition system based on digital metering technology, used to run a data acquisition method based on digital metering technology as described in any embodiment of the present invention, including:

[0080] The sensor condition assessment module includes:

[0081] The temperature drift analysis unit is electrically connected to the temperature probe on the sensor housing and is used to collect and store the temperature-zero offset correspondence.

[0082] The response time monitoring unit has a standard signal generator connected to its input and a timer connected to its output to calculate the rate of change of response time.

[0083] The health calculation unit has its input terminals connected to the temperature drift analysis unit, response time monitoring unit, and zero-point calibration unit, respectively, and its output terminal connected to the parameter adjustment module via a data bus.

[0084] The clock synchronization module includes:

[0085] The phase prediction subunit is directly connected to the main controller clock source and has a built-in autoregressive model calculation unit.

[0086] The feedback correction subunit has at least three historical clock offset registers connected to its input and a clock divider connected to its output via a digital-to-analog converter.

[0087] A branch clock generator, the output of which is connected to the trigger interfaces of each sensor through an isolation circuit;

[0088] The environmental coupling compensation module includes:

[0089] The multi-parameter acquisition unit includes a temperature sensor array, a vibration accelerometer, and a voltage monitoring circuit, and is connected to the main controller via a parallel data interface.

[0090] The weight calculation engine connects to the orthogonal experimental database and real-time environmental parameter cache at the input end and to the compensation matrix update unit at the output end.

[0091] An online compensation actuator has its data input end connected to the sensor signal channel and its control end connected to the compensation matrix update unit.

[0092] The data transmission control module includes:

[0093] The priority calculation unit has inputs connected to the data change rate analyzer and the health assessment unit, and integrates a dynamic weight allocator.

[0094] The bandwidth adaptive unit includes a network quality monitoring circuit and a compression code selector. The network quality monitoring circuit obtains the link status in real time through the physical layer interface.

[0095] The redundant control unit has at least three independent network interfaces connected to its output and a priority tag generator connected to its input.

[0096] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0097] In this invention, dynamic sensor calibration ensures measurement accuracy, a phase-compensated clock tree guarantees data acquisition synchronization, and an error compensation model dynamically corrects for the impact of environmental changes on the data. Finally, a data transmission priority and dynamic bandwidth allocation strategy ensures efficient use of network resources, avoiding packet loss or latency issues when dealing with large amounts of data. Overall, this method not only improves system accuracy but also optimizes data processing and transmission efficiency, providing an effective solution for high-precision data acquisition in large-scale, complex environments.

[0098] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0099] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A data acquisition method based on digital metering technology, characterized in that, Perform the following steps in sequence: Step 1: Perform dynamic calibration of the sensor. By continuously collecting temperature drift curves, response time change rate and zero offset from the sensor's historical working data, calculate the coefficient of variation of each parameter to generate a real-time health index. Based on the correlation between this real-time health index and the real-time detected ambient temperature and humidity, dynamically adjust the sensor's reference parameters. Step 2: Construct a phase-compensated clock tree, generate branch clock signals with forward prediction and feedback correction mechanisms based on the main controller clock, and collect physical quantity signals and corresponding sensor housing temperature gradient, power supply voltage fluctuation extreme environmental parameters under synchronous clock control. Step 3: Establish an error compensation model that couples temperature, humidity and vibration, determine the interaction weights of each factor through orthogonal experiment, and update the compensation coefficients online using the recursive least squares method with a forgetting factor. Step 4: Generate transmission priority based on the product of data change rate and health index, dynamically allocate transmission bandwidth based on network quality assessment value calculated based on latency change rate and packet loss rate, and implement preamble-marked compression encoding on high priority data.

2. The data acquisition method based on digital metering technology according to claim 1, characterized in that, In step 1, the temperature drift curve is obtained by recording the correspondence between the sensor housing temperature and the zero-point offset, specifically as follows: In a constant temperature environment, the zero-point offset is recorded at a preset temperature interval to form a reference curve. In actual work, the temperature sensor detects the shell temperature in real time and interpolates the reference curve to obtain the expected offset corresponding to the current temperature. The response time change rate is calculated by measuring the time difference from 10% to 90% of the steady-state value of the output value after inputting a standard step signal to the sensor, and the change rate is updated once in each sampling period; the zero-point offset is obtained by collecting the sensor output value in the state without input and eliminating random noise by using the sliding window averaging method.

3. The data acquisition method based on digital metering technology according to claim 2, characterized in that, When dynamically adjusting the baseline parameters in step 1, a compensation coefficient lookup table containing a two-dimensional combination of temperature and humidity is first established. The basic compensation amount under the current environmental conditions is obtained by bilinear interpolation. Then, the health index is converted into a confidence factor in the range of 0 to 1. Finally, the calibration parameters are calculated by multiplying the original parameters by (1 + confidence factor × basic compensation amount). The confidence factor is piecewise linearly converted according to the preset interval of the health index.

4. The data acquisition method based on digital metering technology according to claim 3, characterized in that, The process of constructing the phase-compensated clock tree in step 2 includes: Using the clock frequency of the front controller as a reference, the initial frequency division coefficient is determined based on the maximum sampling rate of the sensor to generate the branch clock. An autoregressive model is used to predict the clock offset of the next cycle, and the model order is determined based on the number of zero-crossing points of the autocorrelation function of the historical offset data. In the feedback correction stage, the actual offset of the last three clock cycles is collected, and the weighted average offset is calculated by assigning decreasing weights according to the time proximity. Finally, the clock jitter value is made to decay exponentially by dynamically adjusting the frequency division coefficient until it stabilizes within the preset allowable range.

5. A data acquisition method based on digital metering technology according to claim 4, characterized in that, In step 2, the environmental parameter acquisition and physical quantity sampling are strictly synchronized. The specific implementation method is as follows: The physical quantity sampling channel and the environmental parameter acquisition channel are triggered simultaneously on the rising edge of the branch clock; the peak hold circuit is used to capture the maximum and minimum voltage values ​​at the moment of sampling for recording the extreme values ​​of power supply voltage fluctuations. The temperature gradient of the outer shell is calculated as a linear rate of change using measurements from three temperature sensors arranged along the sensor axis; the vibration spectrum characteristics are obtained by an accelerometer and the dominant frequency component is extracted using a fast Fourier transform.

6. A data acquisition method based on digital metering technology according to claim 5, characterized in that, The implementation of the orthogonal experimental method in step 3 includes: Three test levels (low, medium, and high) were set up for each of the three dimensions of temperature, humidity, and vibration. The sensor output deviation under each combination of conditions was measured through a full factorial experiment. The main effects and pairwise interaction effects of each factor were separated using analysis of variance. The weight of the main effect was allocated according to the contribution ratio, and the weight of the interaction effect was calculated according to the ratio of the square of the effect value to the sum of the squares of the three factors. The forgetting factor was dynamically adjusted according to the logarithmic function value of the rate of change of environmental parameters. When the rate of change exceeded the threshold, the forgetting factor was automatically increased.

7. A data acquisition method based on digital metering technology according to claim 6, characterized in that, The transmission priority generation process in step 4 specifically includes: The data change rate is normalized by calculating the moving average of the differences between adjacent sampling points and dividing it by the sensor range. The health index is nonlinearly compressed using a piecewise processing method. When the index is below the first threshold, a logarithmic transformation is used, and when it is above the second threshold, a square root transformation is used. The final priority value is obtained by adding the normalized change rate and the transformed health index according to dynamic weights, which are linearly adjusted between 0.3 and 0.7 based on the network congestion level.

8. A data acquisition method based on digital metering technology according to claim 7, characterized in that, The specific implementation of the bandwidth allocation strategy in step 4 includes: The latency change rate is calculated by measuring the standard deviation of the round-trip time by continuously sending test data packets; the packet loss rate is statistically analyzed using a variable window size counting method, with the window duration being an integer multiple of the current data sampling period; the compression algorithm is selected by querying the preset strategy table based on the latency-packet loss rate joint evaluation value, and a hybrid compression mode is activated when the evaluation value is in the critical region, using lossless compression for the data packet header and lossy compression for the data payload.

9. A data acquisition method based on digital metering technology according to claim 8, characterized in that, Also includes: When the data jump between adjacent sampling points exceeds the dynamic threshold determined by the health index, a multi-stage data repair process is initiated. First, the normal data trend is predicted based on the autoregressive moving average model. When the prediction residual exceeds the allowable range, the process switches to associated sensor data for spatial interpolation compensation. Finally, an abnormal event report containing the original data, repaired data, environmental parameters, and health index is generated. The dynamic threshold is calculated by multiplying the base threshold by (1 + health index attenuation coefficient), where the attenuation coefficient is obtained from a preset parameter table according to the sensor type.

10. A data acquisition system based on digital metering technology, used to run the data acquisition method based on digital metering technology according to any one of claims 1-9, characterized in that, include: The sensor condition assessment module includes: The temperature drift analysis unit is electrically connected to the temperature probe on the sensor housing and is used to collect and store the temperature-zero offset correspondence. The response time monitoring unit has a standard signal generator connected to its input and a timer connected to its output to calculate the rate of change of response time. The health calculation unit has its input terminals connected to the temperature drift analysis unit, response time monitoring unit, and zero-point calibration unit, respectively, and its output terminal connected to the parameter adjustment module via a data bus. The clock synchronization module includes: The phase prediction subunit is directly connected to the main controller clock source and has a built-in autoregressive model calculation unit. The feedback correction subunit has at least three historical clock offset registers connected to its input and a clock divider connected to its output via a digital-to-analog converter. A branch clock generator, the output of which is connected to the trigger interfaces of each sensor through an isolation circuit; The environmental coupling compensation module includes: The multi-parameter acquisition unit includes a temperature sensor array, a vibration accelerometer, and a voltage monitoring circuit, and is connected to the main controller via a parallel data interface. The weight calculation engine connects to the orthogonal experimental database and real-time environmental parameter cache at the input end and to the compensation matrix update unit at the output end. An online compensation actuator has its data input end connected to the sensor signal channel and its control end connected to the compensation matrix update unit. The data transmission control module includes: The priority calculation unit has inputs connected to the data change rate analyzer and the health assessment unit, and integrates a dynamic weight allocator. The bandwidth adaptive unit includes a network quality monitoring circuit and a compression code selector. The network quality monitoring circuit obtains the link status in real time through the physical layer interface. The redundant control unit has at least three independent network interfaces connected to its output and a priority tag generator connected to its input.

Images