Intelligent monitor temperature drift correction method and system based on temperature compensation algorithm

Abstract

The embodiment of the invention relates to the technical field of data processing, in particular to an intelligent monitor temperature drift correction method and system based on a temperature compensation algorithm, and the method comprises the steps: collecting the environment temperature data of an environment where an intelligent monitor is located in real time, and obtaining a preset sensor aging factor in the intelligent monitor; based on the environment temperature data and a preset temperature-drift characteristic model, the temperature drift characteristic of the intelligent monitor is extracted, and a nonlinear compensation curve matched with the temperature drift characteristic is dynamically generated; performing iterative optimization processing on compensation parameters of the nonlinear compensation curve by adopting a self-adaptive compensation parameter optimization algorithm, and performing dynamic correction on the compensation parameters by fusing sensor aging factors in the iterative optimization processing process to obtain an optimized compensation parameter set; and performing real-time correction processing on original measurement data of the intelligent monitor according to the optimized compensation parameter set, and outputting a target measurement result after temperature drift noise suppression.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular, to a temperature drift correction method and system for an intelligent monitor based on a temperature compensation algorithm. Background Art

[0002] In the field of industrial monitoring, temperature drift correction is a key technology to ensure the measurement accuracy of sensors. Since temperature changes can cause physical property changes in the sensitive elements of sensors, resulting in drift errors in the output signals, traditional methods mostly use fixed compensation coefficients or linear temperature compensation models for correction. Such technologies usually establish a linear relationship between temperature and drift based on laboratory calibration data, and correct the measurement results by presetting compensation parameters.

[0003] However, existing compensation models mostly adopt ideal linear assumptions and cannot accurately characterize the non-linear drift characteristics of sensors under complex temperature gradients, resulting in an increase in compensation residuals in high-low temperature alternating scenarios. Moreover, the compensation parameters of existing technologies are used in a fixed manner for a long time, without considering the time-varying problem of drift characteristics caused by component aging during the actual operation of sensors. As the service time extends, the initial calibration parameters gradually mismatch with the actual characteristics. In response to this situation, current solutions mostly rely on regular manual calibration and maintenance, which are difficult to meet the requirements of continuous and stable measurement in industrial fields. It can be seen that existing technologies have defects such as fast attenuation of compensation accuracy and poor long-term stability. How to construct a collaborative compensation mechanism that dynamically adapts to environmental temperature changes and device aging processes is a technical barrier that urgently needs to be broken through at present. Summary of the Invention

[0004] In order to at least overcome the above deficiencies in the prior art, one of the objectives of the present invention is to provide a temperature drift correction method and system for an intelligent monitor based on a temperature compensation algorithm.

[0005] An embodiment of the present invention provides a temperature drift correction method for an intelligent monitor based on a temperature compensation algorithm, including: Collecting real-time environmental temperature data of the environment where the intelligent monitor is located, and obtaining the sensor aging factor preset in the intelligent monitor; Based on the environmental temperature data and the preset temperature-drift characteristic model, extracting the temperature drift characteristics of the intelligent monitor, and dynamically generating a non-linear compensation curve matching the temperature drift characteristics; Using an adaptive compensation parameter optimization algorithm to perform iterative optimization processing on the compensation parameters of the non-linear compensation curve, and dynamically correcting the compensation parameters by integrating the sensor aging factor during the iterative optimization processing to obtain an optimized set of compensation parameters; Performing real-time correction processing on the original measurement data of the intelligent monitor according to the optimized set of compensation parameters, and outputting a target measurement result after suppressing temperature drift noise.

[0006] In one implementation, extracting the temperature drift characteristics of the intelligent monitor based on the environmental temperature data and a preset temperature-drift characteristic model includes: Dividing the environmental temperature data into multiple temperature data segments according to a preset time window, and performing temperature fluctuation trend analysis on each temperature data segment to obtain the temperature change rate and temperature fluctuation amplitude of each temperature data segment; Invoking the temperature-drift characteristic model to perform non-linear correlation mapping processing on the temperature change rate and temperature fluctuation amplitude, and generating the temperature drift characteristics of the intelligent monitor; the temperature drift characteristics include a first mapping relationship between the temperature change rate and the drift amount, and a second mapping relationship between the temperature fluctuation amplitude and the drift amount.

[0007] In one implementation, dynamically generating a non-linear compensation curve matching the temperature drift characteristics includes: Mapping the temperature change rate to a corresponding first drift component according to the first mapping relationship, and mapping the temperature fluctuation amplitude to a corresponding second drift component according to the second mapping relationship; wherein, the first drift component and the second drift component are dimensionless relative change rates obtained after normalization processing; Inputting the first drift component and the second drift component into a preset compensation curve generation function, and obtaining initial compensation parameters through non-linear weighted fusion calculation; Dynamically adjusting the initial compensation parameters according to the model parameters in the temperature-drift characteristic model to generate adjusted compensation parameters; Constructing a piecewise continuous non-linear compensation curve based on the adjusted compensation parameters, and the non-linear compensation curve is used to provide different compensation amounts in different temperature fluctuation intervals.

[0008] In one implementation, the update process of the preset compensation curve generation function includes: Monitoring the average correction error of the non-linear compensation curve within a preset time period, and determining whether the average correction error exceeds a dynamic adjustment threshold; If it exceeds, adjusting the coefficient of the non-linear relationship expression in the compensation curve generation function according to the distribution characteristics of the average correction error; Reapplying the adjusted compensation curve generation function to the subsequent processing of temperature drift characteristics to achieve the adaptive update of the non-linear compensation curve.

[0009] In one implementation, obtaining the preset sensor aging factor in the intelligent monitor includes: Obtain the historical working duration of the intelligent monitor and the sensor aging weight allocation table; wherein, the sensor aging weight allocation table includes aging impact weights corresponding to different working duration intervals; Match the corresponding aging impact weight from the sensor aging weight allocation table according to the historical working duration, and perform weighted processing on the reference aging factor based on the aging impact weight to generate a dynamically updated sensor aging factor.

[0010] In one implementation, the generation process of the sensor aging weight allocation table includes: Obtain the historical data of multiple reference intelligent monitors during the aging test stage, where the historical data includes sensor performance attenuation indicators at different working durations; According to the correlation between the sensor performance attenuation indicator and the temperature drift correction error, divide the impact levels corresponding to different working duration intervals; Based on the impact levels, allocate corresponding aging impact weights to each working duration interval to generate the sensor aging weight allocation table.

[0011] In one implementation, the use of the adaptive compensation parameter optimization algorithm to perform iterative optimization processing on the compensation parameters of the non-linear compensation curve includes: Obtain the temperature drift correction error in the current iteration cycle, and compare the temperature drift correction error with a preset error threshold; If the temperature drift correction error is greater than the preset error threshold, adjust the parameter values in the compensation parameter set according to the error gradient direction to generate an updated compensation parameter set; Re-enter the updated compensation parameter set into the non-linear compensation curve for parameter verification until the temperature drift correction error is less than or equal to the preset error threshold.

[0012] In one implementation, the fusion of the sensor aging factor to dynamically correct the compensation parameters during the iterative optimization process includes: Divide a preset sensor aging factor value interval according to the numerical value of the sensor aging factor, and configure corresponding compensation parameter correction weights for each sensor aging factor value interval; Extract the offset compensation parameter and the temperature drift slope compensation parameter in the current iteration cycle from the compensation parameter set, and decompose the offset compensation parameter and the temperature drift slope compensation parameter into independent parameter components according to the parameter type; wherein, the independent parameter component is a dimensionless standardized parameter component obtained after the dimension reduction process; Match the corresponding compensation parameter correction weight according to the numerical range of the sensor aging factor, and assign the compensation parameter correction weight to each independent parameter component; Based on the parameter change direction of each independent parameter component, perform a directional weighting operation on the assigned compensation parameter correction weight to generate an aging correction component corresponding to each independent parameter component; Superimpose the aging correction component onto the corresponding independent parameter component to generate a corrected offset compensation parameter component and a corrected temperature drift slope compensation parameter component; Recombine the corrected offset compensation parameter component and the corrected temperature drift slope compensation parameter component according to the parameter type to generate a dynamically corrected compensation parameter set; Input the dynamically corrected compensation parameter set into the nonlinear compensation curve for parameter verification to obtain the verified temperature drift correction error; According to the error distribution characteristics of the verified temperature drift correction error, inversely adjust the compensation parameter correction weight corresponding to the sensor aging factor numerical range to generate an updated compensation parameter correction weight table; Based on the updated compensation parameter correction weight table, perform dynamic correction weight assignment on the independent parameter components in the next iteration cycle to generate an optimized compensation parameter set.

[0013] In one implementation, the real-time correction process of the original measurement data of the intelligent monitor according to the optimized compensation parameter set and outputting the target measurement result after suppressing the temperature drift noise includes: Divide the original measurement data into multiple time-synchronized measurement data segments according to the acquisition timestamp, and the time range of each measurement data segment corresponds one-to-one with the time range of the environmental temperature data segment; According to the time range of the environmental temperature data segment, match the corresponding compensation parameter group from the optimized compensation parameter set, and assign the compensation parameter group to the time-synchronized measurement data segment; Based on the offset correction coefficient and the temperature drift slope in the compensation parameter group, perform a linear compensation operation on each measurement value in the time-synchronized measurement data segment to generate a compensated measurement data sequence; Perform a sliding window mean filtering process on the compensated measurement data sequence, calculate the average value of the measurement data in each window according to the preset window length and overlapping step, and generate a filtered measurement data sequence; Merge the filtered measurement data sequence in timestamp order, remove the measurement data corresponding to the repeated timestamps, and generate the target measurement result after suppressing the temperature drift noise.

[0014] In a non-limiting implementation, after the target measurement result with temperature drift noise suppressed in the output, the following steps are further included: Retrieve from the historical database a historical temperature drift correction data set that matches the current environmental temperature data, and extract the corresponding historical compensation parameter set and historical correction result; Compare the historical compensation parameter set with the optimized compensation parameter set in terms of parameter differences to identify abnormal compensation parameter components that deviate from the preset reference range; Based on the residual distribution characteristics of the historical correction result and the target measurement result, construct a parameter error transfer link to locate the error source direction of the abnormal compensation parameter components; Perform a gradient descent compensation operation on the abnormal compensation parameter components according to the error source direction to generate a backward-corrected compensation parameter component; Re-fit the backward-corrected compensation parameter component with the current environmental temperature data to generate a feedback-corrected compensation parameter set and update it to the non-linear compensation curve.

[0015] In a non-limiting implementation, after the target measurement result with temperature drift noise suppressed in the output, the following steps are further included: Real-time monitor the instantaneous change gradient of subsequent environmental temperature data, and generate a temperature transition event identifier when the instantaneous change gradient exceeds the mutation threshold; Extract the temperature mean difference between the pre-transition steady state interval and the post-transition steady state interval corresponding to the temperature transition event identifier, and determine the product result of the temperature mean difference and the preset transition temperature difference compensation coefficient; Dynamically adjust the temperature drift slope parameter in the optimized compensation parameter set according to the product result to generate a pre-adjusted compensation parameter group; Load the pre-adjusted compensation parameter group in the pre-transition stage of the post-transition steady state interval of the temperature transition, and collect the measurement data fluctuation characteristics in real time during the transition stage; Fine-tune the pre-adjusted compensation parameter group based on the measurement data fluctuation characteristics to generate a steady state compensation parameter group and lock it to the non-linear compensation curve.

[0016] An embodiment of the present invention further provides a temperature drift correction system for an intelligent monitor, including a processor, a memory, and a bus connected to the processor; wherein, the processor and the memory complete communication with each other through the bus; the processor is used to call program instructions in the memory to execute the above-mentioned intelligent monitor temperature drift correction method based on the temperature compensation algorithm.

[0017] An embodiment of the present invention also provides a computer-readable storage medium, on which a program is stored, and when the program is executed by a processor, the above-mentioned temperature drift correction method of the intelligent monitor based on the temperature compensation algorithm is implemented.

[0018] A temperature drift correction method and system for an intelligent monitor based on a temperature compensation algorithm provided by an embodiment of the present invention achieve precise suppression of temperature drift noise and improvement of long-term stability through a multi-dimensional dynamic compensation mechanism. First, based on the real-time collected ambient temperature data and the preset temperature-drift characteristic model, the temperature drift characteristics of the sensor under the current working conditions are dynamically analyzed, and a non-linear compensation curve adapted to the corresponding temperature distribution is generated, breaking through the linear limitation of the traditional fixed compensation model. Secondly, an adaptive compensation parameter optimization algorithm is adopted. By continuously iteratively optimizing the compensation parameters, the non-linear compensation curve can dynamically track the ambient temperature fluctuation, realizing the adaptive evolution of the compensation strategy. In addition, the sensor aging factor is innovatively incorporated into the parameter optimization process, and the characteristic offset caused by device aging is compensated synchronously during iterative correction, constructing a cooperative suppression system for double interference of temperature drift and device aging. Through real-time correction processing, the optimized compensation parameter set is dynamically fused with the original measurement data, effectively eliminating the non-linear interference of temperature drift noise on the measurement result, and at the same time overcoming the problem of long-term compensation failure caused by ignoring device aging in the traditional method. With such a design, a cooperative compensation mechanism that dynamically adapts to environmental temperature changes and device aging processes can be constructed, significantly improving the measurement accuracy stability in a complex temperature field environment, avoiding the compensation lag caused by parameter solidification in conventional temperature drift correction technologies, and realizing self-adaptive compensation calibration throughout the life cycle. Description of the Drawings

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required to be used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0020] Figure 1 It is a flowchart of a temperature drift correction method for an intelligent monitor based on a temperature compensation algorithm provided by an embodiment of the present invention.

[0021] Figure 2 It is a block diagram of a temperature drift correction system for an intelligent monitor provided by an embodiment of the present invention.

[0022] Icons: 100 - Temperature drift correction system for intelligent monitor; 101 - Processor; 102 - Memory; 103 - Bus. Detailed Embodiments

[0023] Exemplary embodiments disclosed in the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present invention can be more thoroughly understood and the scope of the present invention can be completely conveyed to those skilled in the art.

[0024] To better understand the above technical solution, the technical solution of the present invention will be described in detail below with reference to the drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. Without conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0025] Figure 1 FIG. is a flowchart of a temperature drift correction method for an intelligent monitor based on a temperature compensation algorithm according to an embodiment of the present invention, which is applied to an intelligent monitor temperature drift correction system and includes steps 101 to 104.

[0026] Step 101: Real-time collect environmental temperature data of the environment where the intelligent monitor is located, and obtain the sensor aging factor preset in the intelligent monitor.

[0027] In an embodiment of the present invention, the intelligent monitor can be applied to a device status monitoring scenario in an industrial production workshop. The operation of the devices in the industrial production workshop is affected by the environmental temperature. Therefore, it is necessary to collect the environmental temperature in real time. The intelligent monitor is equipped with a high-precision temperature sensor and will collect environmental temperature data every 10 seconds. For example, starting to collect at a certain moment, the collected environmental temperature data is 25°C, 25.2°C, 25.5°C, etc. These data constitute an environmental temperature data sequence.

[0028] At the same time, information related to the sensor aging factor is preset in the intelligent monitor. In an alternative embodiment, obtaining the sensor aging factor preset in the intelligent monitor includes: Step 1011: Obtain the historical working hours of the intelligent monitor and the sensor aging weight allocation table; wherein, the sensor aging weight allocation table includes the aging influence weights corresponding to different working hour intervals, and the generation process of the sensor aging weight allocation table includes: obtaining the historical data of multiple reference intelligent monitors during the aging test stage, and the historical data includes the sensor performance attenuation indexes at different working hours; according to the correlation between the sensor performance attenuation index and the temperature drift correction error, dividing the influence levels corresponding to different working hour intervals; based on the influence levels, allocating the corresponding aging influence weights to each working hour interval to generate the sensor aging weight allocation table.

[0029] In the scenario of the above industrial production workshop, the intelligent monitor has been in long-term use, and through the equipment management system, the historical working hour data of the intelligent monitor can be obtained. For example, for the generation of the sensor aging weight allocation table, the R & D personnel selected 10 reference intelligent monitors of the same model as this intelligent monitor for aging tests. During the aging test, the sensor performance attenuation indexes of these reference intelligent monitors at different working hours were recorded. For example, when working for 1000 hours, the measurement accuracy of the sensor of a certain reference intelligent monitor decreased by 0.5%; when working for 2000 hours, the response time of the sensor of another reference intelligent monitor increased by 10%, etc. Then, based on a large amount of experimental data and analysis, it was found that the larger the sensor performance attenuation index, the larger the temperature drift correction error. Based on this correlation, different working hour intervals were divided, such as 0 - 1000 hours, 1001 - 2000 hours, 2001 - 3000 hours, etc., and according to the influence degree of the sensor performance attenuation on the temperature drift correction error within each interval, the corresponding influence levels were determined, such as low, medium, high. Finally, the corresponding aging influence weights were allocated to each working hour interval. For example, the aging influence weight corresponding to 0 - 1000 hours is 0.1, the aging influence weight corresponding to 1001 - 2000 hours is 0.3, the aging influence weight corresponding to 2001 - 3000 hours is 0.5, etc., thus generating the sensor aging weight allocation table.

[0030] Optionally, in the actual application of the intelligent monitor, the sensor aging weight allocation table is crucial for accurately evaluating and compensating the influence of sensor aging on the measurement results. It details the aging influence weights corresponding to different working hour intervals, and these weights provide a key basis for accurately correcting the sensor data under complex working conditions in the future.

[0031] The generation of the sensor aging weight distribution table first requires selecting multiple reference intelligent monitors of the same model as the actually used intelligent monitor for aging tests. Selecting devices of the same model is to ensure the relevance and accuracy of the test results to the actual application to the greatest extent. In the aging test stage, these reference intelligent monitors are monitored for a long time, and the sensor performance degradation indicators at different working durations are recorded in detail. These indicators cover multiple aspects. For example, the decrease in the measurement accuracy of the sensor may be manifested as the deviation between the measured value and the true value gradually increasing with the increase in the working duration; there is also the problem of the extension of the sensor response time, that is, the time from the input signal to the sensor giving an effective output signal becomes longer. For example, when working for 1000 hours, the measurement accuracy of the sensor of a certain reference intelligent monitor decreased by 0.5%; when working for 2000 hours, the response time of the sensor of another reference intelligent monitor extended by 10%, etc.

[0032] After obtaining the sensor performance degradation index data, the internal correlation between the sensor performance degradation index and the temperature drift correction error can be mined: the larger the sensor performance degradation index, the larger the temperature drift correction error. Based on this, the influence levels can be divided for different working duration intervals. The division process comprehensively considers the influence degree of the sensor performance degradation on the temperature drift correction error, and divides the working duration intervals into multiple levels. For example, 0 - 1000 hours may correspond to a low influence level because the sensor performance degradation is relatively small at this stage and the influence on the temperature drift correction error is also weak; 1001 - 2000 hours corresponds to a medium influence level. At this time, the sensor performance degradation has intensified and the influence on the temperature drift correction error is more obvious; 2001 - 3000 hours and above correspond to a high influence level. As the working duration further increases, the sensor performance degradation is severe and the influence on the temperature drift correction error becomes more prominent.

[0033] Finally, according to the influence level corresponding to each working duration interval, the corresponding aging influence weight is assigned to it. The principle of weight assignment is that the higher the influence level, the greater the weight. For example, the aging influence weight corresponding to 0 - 1000 hours may be 0.1, which means that within this working duration interval, the influence of sensor aging on temperature drift correction is relatively small; the aging influence weight corresponding to 1001 - 2000 hours is 0.3, indicating that the influence degree of sensor aging on temperature drift correction increases at this stage; the aging influence weight corresponding to 2001 - 3000 hours is 0.5, reflecting that the sensor aging has a greater influence on temperature drift correction within this interval. Thus, a comprehensive and accurate sensor aging weight distribution table can be generated.

[0034] Step 1012: Match the corresponding aging impact weight from the sensor aging weight allocation table according to the historical working duration, and perform weighted processing on the benchmark aging factor based on the aging impact weight to generate a dynamically updated sensor aging factor.

[0035] For example, in the embodiment of the present invention, the historical working duration of the intelligent monitor is 5000 hours. By looking up the sensor aging weight allocation table, it is found that it is in the working duration range of "4001 - 5000 hours", and the corresponding aging impact weight for this range is 0.8. For instance, if the benchmark aging factor is 0.5, then through weighted processing, that is, multiplying the aging impact weight by the benchmark aging factor, the dynamically updated sensor aging factor is 0.8 × 0.5 = 0.4.

[0036] Step 102: Based on the environmental temperature data and a preset temperature-drift characteristic model, extract the temperature drift characteristic of the intelligent monitor, and dynamically generate a non-linear compensation curve matching the temperature drift characteristic.

[0037] In the scenario of an industrial production workshop, the environmental temperature data collected by the intelligent monitor is constantly changing. In an optional embodiment, based on the environmental temperature data and a preset temperature-drift characteristic model, extracting the temperature drift characteristic of the intelligent monitor includes: Step 1021: Divide the environmental temperature data into multiple temperature data segments according to a preset time window, and perform temperature fluctuation trend analysis on each temperature data segment to obtain the temperature change rate and temperature fluctuation amplitude of each temperature data segment.

[0038] If the preset time window is set to 60 seconds, the collected environmental temperature data will be divided into multiple 60-second temperature data segments. For example, the data collected in the first temperature data segment is 25°C, 25.2°C, 25.5°C, 25.8°C, 26°C. By calculating the temperature change rate, subtracting the first temperature value from the last temperature value in this data segment and then dividing by the time interval of 60 seconds, that is, (26 - 25) ÷ 60 ≈ 0.0167°C / s, the temperature change rate of this temperature data segment is obtained. The temperature fluctuation amplitude is the difference between the highest temperature value and the lowest temperature value in this data segment, that is, 26 - 25 = 1°C, to obtain the temperature fluctuation amplitude. Similar calculations are performed on each such divided temperature data segment to obtain a series of temperature change rates and temperature fluctuation amplitudes.

[0039] Step 1022: Invoke the temperature-drift characteristic model to perform non-linear correlation mapping processing on the temperature change rate and temperature fluctuation amplitude to generate the temperature drift characteristic of the intelligent monitor; the temperature drift characteristic includes a first mapping relationship between the temperature change rate and the drift amount, and a second mapping relationship between the temperature fluctuation amplitude and the drift amount.

[0040] Optionally, the temperature-drift characteristic model is established based on a large amount of experimental data and analysis. In the embodiments of the present invention, the temperature change rate and the temperature fluctuation amplitude of each calculated temperature data segment are input into the model. For example, for a data segment with a temperature change rate of 0.0167 °C / second and a temperature fluctuation amplitude of 1 °C, through the non-linear correlation mapping process of the model, a first mapping relationship between the temperature change rate and the drift amount is obtained. For example, under this model, when the temperature change rate is 0.0167 °C / second, the corresponding drift amount is 0.05; at the same time, a second mapping relationship between the temperature fluctuation amplitude and the drift amount is obtained. When the temperature fluctuation amplitude is 1 °C, the corresponding drift amount is 0.1. These mapping relationships together constitute the temperature-drift characteristic of the intelligent monitor.

[0041] In an alternative embodiment, the dynamically generating a non-linear compensation curve matching the temperature-drift characteristic includes: Step 1023: According to the first mapping relationship, map the temperature change rate to a corresponding first drift component, and according to the second mapping relationship, map the temperature fluctuation amplitude to a corresponding second drift component; wherein, the first drift component and the second drift component are dimensionless relative change rates obtained after normalization processing.

[0042] For the previously obtained temperature change rate of 0.0167 °C / second, the corresponding drift amount is 0.05. For example, the maximum value of this temperature change rate during the entire monitoring process is 0.1 °C / second, and the minimum value is 0 °C / second. Through normalization processing, the first drift component = (0.0167 - 0) ÷ (0.1 - 0) = 0.167. For the temperature fluctuation amplitude of 1 °C, the corresponding drift amount is 0.1. For example, the maximum value of the temperature fluctuation amplitude is 3 °C, and the minimum value is 0 °C. Then the second drift component = (1 - 0) ÷ (3 - 0) = 0.333. Thus, the dimensionless first drift component and second drift component obtained after normalization processing can be obtained.

[0043] Step 1024: Input the first drift component and the second drift component into a preset compensation curve generation function, and calculate an initial compensation parameter through non-linear weighted fusion.

[0044] Among them, the preset compensation curve generation function takes into account the different influence degrees of the first drift component and the second drift component. The weight of the first drift component can be set to 0.4, and the weight of the second drift component can be set to 0.6. In the non-linear weighted fusion calculation, for the first drift component 0.167 and the second drift component 0.333, first multiply them by their respective weights, that is, 0.167×0.4 = 0.0668, 0.333×0.6 = 0.1998, and then add these two results, 0.0668 + 0.1998 = 0.2666, to obtain the initial compensation parameter.

[0045] Among them, the update process of the preset compensation curve generation function includes: monitoring the average correction error of the non-linear compensation curve within a preset time period, and judging whether the average correction error exceeds the dynamic adjustment threshold; if it exceeds, adjust the coefficient of the non-linear relationship expression in the compensation curve generation function according to the distribution characteristics of the average correction error; apply the adjusted compensation curve generation function to the subsequent processing of the temperature drift characteristics again to realize the adaptive update of the non-linear compensation curve.

[0046] In the embodiment of the present invention, the preset time period is set to 1 hour. The average correction error of the non-linear compensation curve within this time period is calculated once per hour. For example, within a certain hour, after correcting the measurement data of the intelligent monitor through the non-linear compensation curve, the calculated average correction error is 0.2. For example, the dynamic adjustment threshold is 0.15. Since 0.2 exceeds 0.15, at this time, analyze the distribution characteristics of the average correction error and find that the error is mainly concentrated in the region with a large temperature change rate. Then, according to this feature, adjust the coefficient of the non-linear relationship expression related to the temperature change rate in the compensation curve generation function. For example, increase the coefficient related to the first drift component to enhance the compensation ability for the case of a large temperature change rate. After adjustment, apply the new compensation curve generation function to the subsequent processing of the temperature drift characteristics.

[0047] More specifically, the update process of the preset compensation curve generation function is a dynamic and adaptive mechanism, aiming to ensure that the non-linear compensation curve can always accurately adapt to the temperature drift characteristics of the intelligent monitor in different environments, thereby improving the accuracy and stability of the measurement data.

[0048] First, the system continuously and precisely monitors the average calibration error of the non-linear compensation curve within a preset time period. The setting of the preset time period needs to comprehensively consider various factors, such as the change frequency of measurement data, the stability of ambient temperature, and the measurement accuracy requirements of actual applications, etc. In the embodiment of the present invention, the preset time period is set to 1 hour (which can not only capture the change of the calibration effect of the compensation curve in time but also avoid calculating and adjusting too frequently). Every hour, the system automatically collects all the error data after the non-linear compensation curve corrects the measurement data of the intelligent monitor within this time period, and calculates the average calibration error. This average calibration error reflects the average deviation degree between the measurement data and the true value after the non-linear compensation curve compensates for the temperature drift within this time period.

[0049] Next, the calculated average calibration error is compared with the dynamic adjustment threshold for judgment. The dynamic adjustment threshold is not a fixed value. It will be dynamically adjusted according to factors such as the usage environment of the intelligent monitor, measurement requirements, and long-term monitoring data feedback, etc. For example, in an industrial production scenario with high measurement accuracy requirements, the dynamic adjustment threshold may be set relatively low to ensure the high accuracy of measurement data; while in some scenarios with relatively loose accuracy requirements, the threshold can be appropriately increased. If the average calibration error exceeds the dynamic adjustment threshold, it indicates that the non-linear compensation curve generated by the current compensation curve generation function may not effectively compensate for the temperature drift, and the error of the measurement data exceeds the acceptable range, and the compensation curve generation function needs to be adjusted.

[0050] When the average calibration error exceeds the dynamic adjustment threshold, the system analyzes the distribution characteristics of the average calibration error. For example, by classifying and statistically analyzing the error data according to factors such as the temperature change rate and the temperature fluctuation amplitude, it is found in which regions or situations the errors are mainly concentrated. If it is determined that the errors are mainly concentrated in the region with a large temperature change rate, it indicates that in this case, the current compensation curve generation function is insufficient in compensating for the temperature drift generated during rapid temperature changes; or if the errors are concentrated in the region with a large temperature fluctuation amplitude, it means that the compensation curve has poor effect in dealing with large-amplitude temperature fluctuations.

[0051] According to the distribution characteristics of the average calibration error, the system will specifically adjust the coefficients of the non-linear relationship expression in the compensation curve generation function. These coefficients determine the sensitivity and compensation ability of the compensation curve generation function to different factors. For example, if it is found that the errors are mainly concentrated in the region with a large temperature change rate, then the coefficient of the non-linear relationship expression related to the temperature change rate will be increased, so as to enhance the compensation ability of the compensation curve generation function for the situation with a large temperature change rate, making the generated non-linear compensation curve more effectively correct the measurement data and reduce the error when facing rapid temperature changes.

[0052] Furthermore, the adjusted compensation curve generation function will be reapplied to the subsequent processing of temperature drift characteristics. In the subsequent processing process, the new compensation curve generation function will generate a nonlinear compensation curve that is more in line with the actual situation based on the real-time collected ambient temperature data and the temperature drift characteristics obtained by the latest analysis. This new nonlinear compensation curve will provide more accurate differentiated compensation amounts in different temperature fluctuation ranges, thereby realizing the adaptive update of the nonlinear compensation curve. For example, in a range where the temperature fluctuation is relatively gentle, the compensation curve may use a relatively gentle function relationship for compensation; while in a range where the temperature fluctuates violently, the compensation curve will use a more aggressive function relationship based on the adjusted coefficient to enhance the compensation effect. Through the above-mentioned adaptive update mechanism, the nonlinear compensation curve can better adapt to the changing ambient temperature, effectively suppress the impact of temperature drift on the measurement data, and ensure that the intelligent monitor can always provide accurate and stable measurement results in a complex temperature environment, meeting the strict requirements for measurement accuracy in actual application scenarios such as industrial production.

[0053] Step 1025: dynamically adjusting the initial compensation parameters according to the model parameters in the temperature-drift characteristic model to generate adjusted compensation parameters.

[0054] There are some fixed parameters in the temperature-drift characteristic model, such as the scale factor in the model. In the embodiment of the present invention, for example, a scale factor in the model is 1.2. The initial compensation parameter 0.2666 is multiplied by the scale factor, that is, 0.2666×1.2=0.31992, to obtain the adjusted compensation parameter.

[0055] Step 1026: constructing a piecewise continuous nonlinear compensation curve based on the adjusted compensation parameters, wherein the nonlinear compensation curve is used to provide differentiated compensation amounts within different temperature fluctuation ranges.

[0056] In the ambient temperature range of industrial production workshops, different intervals are divided according to temperature fluctuations, such as 20-25°C, 25-30°C, etc. Based on the adjusted compensation parameter 0.31992, combined with the characteristics of different temperature intervals, a piecewise continuous nonlinear compensation curve is constructed. For example, in the 20-25°C interval, according to the preset functional relationship (such as a quadratic function relationship), the compensation amount corresponding to different temperature points in the interval is calculated based on the compensation parameter; in the 25-30°C interval, another functional relationship (such as an exponential function relationship) is used to calculate the compensation amount, thereby constructing a nonlinear compensation curve that provides differentiated compensation amounts in different temperature fluctuation intervals.

[0057] Step 103: Use the adaptive compensation parameter optimization algorithm to iteratively optimize the compensation parameters of the non - linear compensation curve, and dynamically correct the compensation parameters by integrating the sensor aging factor during the iterative optimization process to obtain an optimized set of compensation parameters.

[0058] In this industrial monitoring scenario, the adaptive compensation parameter optimization algorithm continuously adjusts the compensation parameters. In an optional embodiment, using the adaptive compensation parameter optimization algorithm to iteratively optimize the compensation parameters of the non - linear compensation curve includes: Step 1031: Obtain the temperature drift correction error in the current iteration cycle, and compare the temperature drift correction error with a preset error threshold.

[0059] For example, the current iteration cycle is the 5th iteration. In this cycle, after correcting the measurement data of the intelligent monitor through the non - linear compensation curve, the calculated temperature drift correction error is 0.18. The preset error threshold is 0.15.

[0060] Step 1032: If the temperature drift correction error is greater than the preset error threshold, adjust the parameter values in the compensation parameter set according to the error gradient direction to generate an updated set of compensation parameters.

[0061] Since 0.18 is greater than 0.15, according to the error gradient direction, that is, the direction in which the error increases, adjust the parameter values in the compensation parameter set. For example, there are parameters a, b, c, etc. in the compensation parameter set. By analyzing the error gradient, it is found that parameter a has a greater impact on the error, and a needs to increase when the error increases. Another example is that the original value of a is 0.5, and according to the error gradient calculation, the value of a is increased to 0.6, thus generating an updated set of compensation parameters.

[0062] Step 1033: Re - input the updated set of compensation parameters into the non - linear compensation curve for parameter verification until the temperature drift correction error is less than or equal to the preset error threshold.

[0063] Input the updated set of compensation parameters into the non - linear compensation curve, and correct the measurement data of the intelligent monitor again. Calculate the new temperature drift correction error. If the new error is still greater than the preset error threshold, continue to adjust the parameters according to the above steps until the temperature drift correction error is less than or equal to 0.15.

[0064] In an optional embodiment, during the iterative optimization process, integrating the sensor aging factor to dynamically correct the compensation parameters to obtain an optimized set of compensation parameters includes: Step 10341: Divide the preset numerical range of the sensor aging factor according to the value of the sensor aging factor, and configure the corresponding compensation parameter correction weight for each numerical range of the sensor aging factor.

[0065] In the embodiment of the present invention, the sensor aging factor is 0.4. According to the preset setting, the numerical range of the sensor aging factor is divided into 0 - 0.2, 0.2 - 0.4, 0.4 - 0.6, etc. For the range of 0.2 - 0.4, the configured compensation parameter correction weight is 0.3.

[0066] Step 10342: Extract the offset compensation parameter and the temperature drift slope compensation parameter in the current iteration period from the compensation parameter set, and decompose the offset compensation parameter and the temperature drift slope compensation parameter into independent parameter components according to the parameter type; wherein, the independent parameter component is a dimensionless standardized parameter component obtained after the dimensionless processing.

[0067] Optionally, there are an offset compensation parameter x and a temperature drift slope compensation parameter y in the compensation parameter set. Decompose x into independent parameter components such as x1, x2, etc. according to the parameter type, and decompose y into independent parameter components such as y1, y2, etc. For example, the original value of x is 10 (unit: a certain physical quantity unit). Through the dimensionless processing, assuming the maximum value of this physical quantity is 100 and the minimum value is 0, then the standardized x1 = (10 - 0) ÷ (100 - 0) = 0.1, and similar processing is performed on x2, etc.; the original value of y is 0.5 (unit: another physical quantity unit), assuming the maximum value of this physical quantity is 1 and the minimum value is 0, then the standardized y1 = (0.5 - 0) ÷ (1 - 0) = 0.5, and similar processing is performed on y2, etc.

[0068] Step 10343: Match the corresponding compensation parameter correction weight according to the numerical range of the sensor aging factor where the sensor aging factor is located, and allocate the compensation parameter correction weight to each independent parameter component.

[0069] Since the sensor aging factor 0.4 is in the range of 0.2 - 0.4, the corresponding compensation parameter correction weight for this range is 0.3. Allocate this weight to each independent parameter component, that is, the correction weight of x1 is 0.3, the correction weight of x2 is 0.3, the correction weight of y1 is 0.3, the correction weight of y2 is 0.3, etc.

[0070] Step 10344: Perform a directional weighting operation on the allocated compensation parameter correction weight based on the parameter change direction of each independent parameter component to generate an aging correction component corresponding to each independent parameter component.

[0071] For example, the parameter change direction of x1 is increasing, and the parameter change direction of y1 is decreasing. For x1, the aging correction component = 0.3 × 1 (the weight for the increasing direction is set to 1) = 0.3; for y1, the aging correction component = 0.3 × (-1) (the weight for the decreasing direction is set to -1) = -0.3.

[0072] Step 10345: Superimpose the aging correction component onto the corresponding independent parameter component to generate a corrected offset compensation parameter component and a corrected temperature drift slope compensation parameter component.

[0073] For x1, the corrected x1 = 0.1 + 0.3 = 0.4; for y1, the corrected y1 = 0.5 - 0.3 = 0.2. And so on, all corrected offset compensation parameter components and corrected temperature drift slope compensation parameter components are obtained.

[0074] Step 10346: Re - combine the corrected offset compensation parameter component and the corrected temperature drift slope compensation parameter component according to the parameter type to generate a dynamically corrected compensation parameter set.

[0075] Re - combine the corrected x1, x2, etc. into offset compensation parameters, and re - combine the corrected y1, y2, etc. into temperature drift slope compensation parameters, thus generating a dynamically corrected compensation parameter set.

[0076] Step 10347: Input the dynamically corrected compensation parameter set into the non - linear compensation curve for parameter verification to obtain the verified temperature drift correction error.

[0077] Input the dynamically corrected compensation parameter set into the non - linear compensation curve, correct the measurement data of the intelligent monitor, and calculate the verified temperature drift correction error, for example, it is 0.12.

[0078] Step 10348: According to the error distribution characteristics of the verified temperature drift correction error, inversely adjust the compensation parameter correction weight corresponding to the numerical range of the sensor aging factor to generate an updated compensation parameter correction weight table.

[0079] Analyze the error distribution characteristics of the verified temperature drift correction error 0.12, and find that the adjustment effect of some parameters is good, while the adjustment effect of some other parameters is not good. According to this situation, inversely adjust the compensation parameter correction weight corresponding to the numerical range of the sensor aging factor. For example, for the range of 0.2 - 0.4, adjust the compensation parameter correction weight from 0.3 to 0.4 to generate an updated compensation parameter correction weight table.

[0080] Step 10349: Dynamically correct the weight distribution of the independent parameter components in the next iteration cycle based on the updated compensation parameter correction weight table to generate an optimized set of compensation parameters.

[0081] In the next iteration cycle, according to the updated compensation parameter correction weight table, reassign the correction weights to the independent parameter components, continue with the parameter adjustment, and finally generate an optimized set of compensation parameters.

[0082] Step 104: Perform real-time calibration processing on the original measurement data of the intelligent monitor according to the optimized set of compensation parameters, and output the target measurement result after suppressing the temperature drift noise.

[0083] In an industrial production workshop scenario, the intelligent monitor continuously collects various data of the equipment, and these original measurement data are affected by temperature drift. In an optional embodiment, performing real-time calibration processing on the original measurement data of the intelligent monitor according to the optimized set of compensation parameters and outputting the target measurement result after suppressing the temperature drift noise includes: Step 1041: Divide the original measurement data into multiple time-synchronized measurement data segments according to the acquisition timestamp, and the time range of each measurement data segment corresponds one-to-one with the time range of the environmental temperature data segment.

[0084] The original measurement data collected by the intelligent monitor involves multiple parameters such as the vibration amplitude and current intensity of the equipment. For example, the data collected regarding the vibration amplitude are a series of values, such as 10 (unit: millimeter), 10.2, 10.5, etc., and the acquisition timestamp of each data is recorded. According to the same time window as the division of the environmental temperature data segment (such as 60 seconds), divide these original measurement data into multiple measurement data segments. The first measurement data segment contains the vibration amplitude data 10, 10.2, 10.5 collected within a certain 60 seconds, the second measurement data segment contains the data collected within the next 60 seconds, and so on, and the time range of each measurement data segment corresponds strictly one-to-one with the time range of the previously divided environmental temperature data segment.

[0085] Step 1042: According to the time range of the environmental temperature data segment, match the corresponding compensation parameter group from the optimized set of compensation parameters, and assign the compensation parameter group to the time-synchronized measurement data segment.

[0086] For the environmental temperature data segment within a certain time range, such as the environmental temperature data segment from 10:00:00 - 10:01:00, find the corresponding compensation parameter group from the optimized compensation parameter set according to its time range. For example, the compensation parameter group includes an offset correction coefficient of 0.5 and a temperature drift slope of 0.2. Assign this compensation parameter group to the measurement data segment synchronized with it in time, that is, the data segment of the original measurement data such as vibration amplitude collected from 10:00:00 - 10:01:00.

[0087] Step 1043: Based on the offset correction coefficient and the temperature drift slope in the compensation parameter group, perform a linear compensation operation on each measurement value in the time - synchronized measurement data segment to generate a compensated measurement data sequence.

[0088] For each vibration amplitude measurement value in the measurement data segment assigned the compensation parameter group, such as the measurement value being 10 (unit: millimeter), perform a linear compensation operation. The compensation formula is: Compensated value = Measurement value+Offset correction coefficient+Temperature drift slope×(Current environmental temperature - Reference temperature). For example, if the current environmental temperature is 25 °C and the reference temperature is 20 °C, then for the measurement value of 10, the compensated value = 10 + 0.5+0.2×(25 - 20)=10 + 0.5 + 1 = 11.5 (unit: millimeter). Perform the above operation on each measurement value in this measurement data segment to generate a compensated measurement data sequence, such as 11.5, 11.7, 12, etc.

[0089] Step 1044: Perform a sliding window mean filtering process on the compensated measurement data sequence. According to the preset window length and overlapping step size, calculate the average value of the measurement data within each window to generate a filtered measurement data sequence.

[0090] The preset window length is 5 and the overlapping step size is 2. For the compensated measurement data sequence 11.5, 11.7, 12, 12.2, 12.5, 12.8..., the first window contains 11.5, 11.7, 12, 12.2, 12.5, and calculate its average value as (11.5 + 11.7+12 + 12.2+12.5)÷5 = 12. The second window starts from the second data and contains 11.7, 12, 12.2, 12.5, 12.8, and calculate its average value as (11.7 + 12+12.2+12.5+12.8)÷5 = 12.24. And so on. Through this sliding window mean filtering process, generate a filtered measurement data sequence, such as 12, 12.24, 12.48, etc.

[0091] Step 1045: Merge the filtered measurement data sequence in the order of timestamps, remove the measurement data corresponding to duplicate timestamps, and generate the target measurement result after suppressing the temperature drift noise.

[0092] Merge the filtered measurement data sequence in the order of the acquisition timestamps. During the merging process, if it is found that there is measurement data corresponding to duplicate timestamps, only keep one of them. For example, during the merging process, if it is found that there are two pieces of measurement data with a timestamp of 10:00:30, only keep one of them. After the above processing, the target measurement result after suppressing the temperature drift noise is finally generated, and these results can more accurately reflect the true state of the device, reducing the influence of temperature drift on the measurement data.

[0093] In a non-limiting embodiment, after outputting the target measurement result after suppressing the temperature drift noise, it further includes: Step 201: Retrieve from the historical database the historical temperature drift correction data set that matches the current ambient temperature data, extract the corresponding historical compensation parameter set and historical correction result, compare the parameter differences between the historical compensation parameter set and the optimized compensation parameter set, identify the abnormal compensation parameter components that deviate from the preset reference range, and based on the residual distribution characteristics of the historical correction result and the target measurement result, construct a parameter error transmission link to locate the error source direction of the abnormal compensation parameter components.

[0094] In the industrial production workshop scenario, the historical database stores the temperature drift correction data at different ambient temperatures over a period of time. According to the current ambient temperature data, for example, the current ambient temperature is 25 °C, retrieve from the historical database the historical temperature drift correction data set with the ambient temperature in the range of 24 °C - 26 °C. Extract the historical compensation parameter set from this data set. For example, a certain parameter in the historical compensation parameter set is 0.4, while the corresponding parameter in the optimized compensation parameter set is 0.6. By comparison, it is found that this parameter deviates from the preset reference range, and the preset reference range is set to 0.3 - 0.5. At the same time, extract the historical correction result, compare the historical correction result with the currently output target measurement result, and calculate the residual distribution characteristics. For example, it is found that the residuals at some measurement points are relatively large. By analyzing these residual distributions, combining the historical data and the current compensation parameter situation, construct a parameter error transmission link, and it is found that the error of this abnormal compensation parameter component may come from inaccurate parameter adjustment of the temperature - drift characteristic model in the current temperature range.

[0095] Step 202: Perform gradient descent compensation operation on the abnormal compensation parameter component according to the error source direction, generate a reversely corrected compensation parameter component, refit the reversely corrected compensation parameter component with the current ambient temperature data, generate a feedback corrected compensation parameter set and update it to the non-linear compensation curve.

[0096] According to the previously located error source direction, perform gradient descent compensation operation on the abnormal compensation parameter component. For example, for the offset compensation parameter 0.6, calculate the gradient direction according to the gradient descent algorithm. The gradient direction indicates that this parameter needs to be decreased. With a preset step size (set the step size to 0.1), reduce the parameter to 0.6 - 0.1 = 0.5. Refit this reversely corrected compensation parameter component 0.5 with the current ambient temperature data of 25 °C, considering factors such as the temperature change rate and the temperature fluctuation amplitude, to generate a feedback corrected compensation parameter set. For example, after fitting, a new set of compensation parameters is obtained, and this set of compensation parameters is updated to the non-linear compensation curve, enabling the non-linear compensation curve to better adapt to the actual situation and improve the correction ability for temperature drift.

[0097] In a non-limiting embodiment, after outputting the target measurement result with suppressed temperature drift noise, it further includes: Step 301: Real-time monitor the instantaneous change gradient of subsequent ambient temperature data. When the instantaneous change gradient exceeds the mutation threshold, generate a temperature transition event identifier, extract the temperature mean difference between the pre-transition steady state interval and the post-transition steady state interval corresponding to the temperature transition event identifier, and determine the product result of the temperature mean difference and the preset transition temperature difference compensation coefficient. Dynamically adjust the temperature drift slope parameter in the optimized compensation parameter set according to the product result to generate a pre-adjusted compensation parameter group.

[0098] In an industrial production workshop, continuously and real-time monitor the instantaneous change gradient of environmental temperature data. For example, the environmental temperature data is 25°C, 25.2°C, 25.5°C, 26°C, 28°C, 30°C... Calculate the change gradient between adjacent two temperature values, such as (25.2 - 25) ÷ (measurement interval time), (25.5 - 25.2) ÷ (measurement interval time), etc. For example, the mutation threshold is 0.5°C / minute. When the instantaneous change gradient at a certain moment exceeds this threshold, for example, rapidly rising from 26°C to 28°C and the change gradient exceeds 0.5°C / minute, generate a temperature transition event identifier. Then, extract the temperature mean difference between the pre-transition steady-state interval (for example, 25°C - 26°C) and the post-transition steady-state interval (for example, 28°C - 30°C) corresponding to this temperature transition event identifier. The temperature mean of the pre-transition steady-state interval is (25 + 26) ÷ 2 = 25.5°C, the temperature mean of the post-transition steady-state interval is (28 + 30) ÷ 2 = 29°C, and the temperature mean difference is 29 - 25.5 = 3.5°C. The preset transition temperature compensation coefficient is set to 0.2, and calculate the product result as 3.5 × 0.2 = 0.7. According to this product result, dynamically adjust the temperature drift slope parameter in the optimized compensation parameter set. For example, the original temperature drift slope parameter is 0.3, and after adjustment, it becomes 0.3 + 0.7 = 1, generating a pre-adjusted compensation parameter group.

[0099] Step 302: Load the pre-adjusted compensation parameter group in the pre-transition stage of the post-temperature-transition steady-state interval, and real-time collect the fluctuation characteristics of the measurement data in the transition stage. Based on the fluctuation characteristics of the measurement data, finely adjust the pre-adjusted compensation parameter group to generate a steady-state compensation parameter group and lock it to the non-linear compensation curve.

[0100] In the pre-transition stage of the post-temperature-transition steady-state interval (such as 28°C - 30°C, taking the stage from 26°C to 28°C as an example), load the pre-adjusted compensation parameter group. In this transition stage, real-time collect the fluctuation characteristics of the measurement data, such as collecting the fluctuation conditions of measurement data such as the vibration amplitude of the device and the current intensity. For another example, it is found that the fluctuation of the vibration amplitude measurement data is relatively large. Analyze this fluctuation characteristic and find that the current pre-adjusted compensation parameter group has an unsatisfactory suppression effect on the fluctuation. According to the analysis result, finely adjust the pre-adjusted compensation parameter group. For example, adjust the offset correction coefficient from 0.5 to 0.6. After fine adjustment, generate a steady-state compensation parameter group, and lock this steady-state compensation parameter group to the non-linear compensation curve, so that the non-linear compensation curve can better correct the measurement data in the new temperature transition situation, further improve the accuracy and stability of the measurement data of the intelligent monitor, effectively suppress the influence of temperature drift noise on the measurement result, and ensure the reliability of the equipment status monitoring in the industrial production workshop.

[0101] In the embodiments of the present invention, a precise suppression of temperature drift noise and an improvement in long-term stability are achieved through a multi-dimensional dynamic compensation mechanism. First, based on the ambient temperature data collected in real time and the preset temperature-drift characteristic model, the temperature-drift characteristics of the sensor under the current working conditions are dynamically analyzed to generate a non-linear compensation curve adapted to the corresponding temperature distribution, breaking through the linear limitation of the traditional fixed compensation model. Second, an adaptive compensation parameter optimization algorithm is adopted. By continuously iteratively optimizing the compensation parameters, the non-linear compensation curve can dynamically track the ambient temperature fluctuations, realizing the adaptive evolution of the compensation strategy. In addition, the sensor aging factor is innovatively incorporated into the parameter optimization process, and the characteristic offset caused by device aging is compensated synchronously during iterative correction, constructing a collaborative suppression system for dual interferences of temperature drift and device aging. Through real-time calibration processing, the optimized set of compensation parameters is dynamically fused with the original measurement data, effectively eliminating the non-linear interference of temperature drift noise on the measurement results and overcoming the problem of long-term compensation failure caused by ignoring device aging in traditional methods. With such a design, a collaborative compensation mechanism that dynamically adapts to environmental temperature changes and device aging processes can be constructed, significantly improving the measurement accuracy stability in complex temperature field environments, avoiding the compensation lag caused by parameter solidification in conventional temperature drift correction technologies, and realizing self-adaptive compensation calibration throughout the entire life cycle.

[0102] Embodiments of the present invention provide a computer-readable storage medium, on which a program is stored, and when the program is executed by a processor, the temperature drift correction method of the intelligent monitor based on the temperature compensation algorithm is implemented.

[0103] Embodiments of the present invention provide a processor, which is used to run a program, and when the program runs, the temperature drift correction method of the intelligent monitor based on the temperature compensation algorithm is executed.

[0104] In the embodiments of the present invention, as Figure 2 shown, the temperature drift correction system 100 of the intelligent monitor includes at least one processor 101, and at least one memory 102 and a bus 103 connected to the processor 101; wherein, the processor 101 and the memory 102 complete communication with each other through the bus 103; the processor 101 is used to call the program instructions in the memory 102 to execute the above-mentioned temperature drift correction method of the intelligent monitor based on the temperature compensation algorithm.

[0105] The present invention is described with reference to the flowcharts and / or block diagrams of methods, intelligent monitor temperature drift correction systems (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or multiple blocks.

[0106] In a typical configuration, an intelligent monitor temperature drift correction system includes one or more processors (CPUs), a memory, and a bus. The intelligent monitor temperature drift correction system may also include an input / output interface, a network interface, etc.

[0107] The memory may include non-permanent memory in the form of computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM). The memory includes at least one storage chip. The memory is an example of computer-readable media.

[0108] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can store information by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory, or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD), or other optical storage, magnetic cassette tapes, magnetic disk storage, or other magnetic storage, computer-readable storage media, or any other non-transmission media that can be used to store information that can be accessed by the intelligent monitor temperature drift correction system. As defined herein, computer-readable media does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0109] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, commodity or computer-readable storage medium comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, commodity or computer-readable storage medium. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, commodity or computer-readable storage medium comprising the element.

[0110] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0111] The above are only embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

[0112] In the technical solutions involved in the above embodiments of the present invention, whether it is for calculating the comparison of multi-dimensional features or constructing composite parameters, when there are problems caused by significant differences in the number of dimensions, dimension units, and semantic meanings of different features, those skilled in the art, based on their professional knowledge and past practical experience, can fully understand that these differences need to be properly handled to make the calculation results accurate and comparable, and to avoid situations such as logical confusion and unclear mathematical meanings.

[0113] Specifically, when faced with features having different numbers of dimensions, in order to accurately calculate the similarity, matching degree or feature distance between different features, those skilled in the art can use various strategies, such as strategies of feature selection, feature extraction, kernel functions, etc. for adaptive processing.

[0114] When dealing with the comparison of multi-dimensional features, in order to achieve comparable alignment of the feature space, those skilled in the art can use a variety of existing general technical means, including but not limited to the following existing technologies: standardization preprocessing, mapping transformation, space projection, etc.

[0115] In the process of constructing composite parameters (such as loss function values), different parameter terms often have different dimensions. Those skilled in the art can adopt existing normalization processing or an adaptive weight allocation mechanism based on distribution characteristics.

[0116] The above general technical means for solving the problems of feature matching and loss balance all belong to the common general knowledge in this field. These technical means have been fully verified and widely used in a large number of practical applications. When facing similar dimension difference problems, those skilled in the art can proficiently and flexibly use these methods for processing.

[0117] The formulas and calculation processes involved in the embodiments of the present invention, whether for multi-dimensional feature comparison or composite loss function construction, strictly follow the principle of dimension correspondence. Each variable in the formula has a clear and definite physical meaning, and its operation logic fully conforms to basic mathematical and physical logics. The operation result is necessarily a reasonable result expected by the present invention. Those skilled in the art are capable of comprehensively using the above general technical means according to specific data situations and business requirements to effectively solve various problems brought about by the number of dimensions, dimension differences, etc. in the multi-dimensional feature comparison calculation and composite loss function construction in the embodiments, ensuring the accuracy, reliability, and feasibility of the technical solution of the present invention.

Claims

1. An intelligent monitor temperature drift correction method based on a temperature compensation algorithm, characterized in that, Including: Collecting in real time the ambient temperature data of the intelligent monitor, and obtaining the sensor aging factor preset in the intelligent monitor; Based on the ambient temperature data and the preset temperature-drift characteristic model, extracting the temperature drift characteristic of the intelligent monitor, and dynamically generating a non-linear compensation curve matching the temperature drift characteristic; Adopting an adaptive compensation parameter optimization algorithm to perform iterative optimization processing on the compensation parameters of the non-linear compensation curve, and dynamically correcting the compensation parameters by integrating the sensor aging factor during the iterative optimization processing to obtain an optimized set of compensation parameters; Performing real-time correction processing on the original measurement data of the intelligent monitor according to the optimized set of compensation parameters, and outputting a target measurement result after suppressing the temperature drift noise.

2. The method according to claim 1, characterized in that, The extracting the temperature drift characteristic of the intelligent monitor based on the ambient temperature data and the preset temperature-drift characteristic model includes: Dividing the ambient temperature data into multiple temperature data segments according to a preset time window, and performing temperature fluctuation trend analysis on each temperature data segment to obtain the temperature change rate and temperature fluctuation amplitude of each temperature data segment; Invoking the temperature-drift characteristic model to perform non-linear correlation mapping processing on the temperature change rate and temperature fluctuation amplitude to generate the temperature drift characteristic of the intelligent monitor; the temperature drift characteristic includes a first mapping relationship between the temperature change rate and the drift amount, and a second mapping relationship between the temperature fluctuation amplitude and the drift amount.

3. The method according to claim 2, characterized in that, The dynamically generating a non-linear compensation curve matching the temperature drift characteristic includes: According to the first mapping relationship, mapping the temperature change rate to a corresponding first drift component, and according to the second mapping relationship, mapping the temperature fluctuation amplitude to a corresponding second drift component; wherein, the first drift component and the second drift component are dimensionless relative change rates obtained after normalization processing; Inputting the first drift component and the second drift component into a preset compensation curve generation function, and obtaining initial compensation parameters through non-linear weighted fusion calculation; Dynamically adjusting the initial compensation parameters according to the model parameters in the temperature-drift characteristic model to generate adjusted compensation parameters; Constructing a piecewise continuous non-linear compensation curve based on the adjusted compensation parameters, and the non-linear compensation curve is used to provide different compensation amounts in different temperature fluctuation intervals.

4. The method according to claim 3, wherein The update process of the preset compensation curve generation function includes: Monitoring the average correction error of the non-linear compensation curve within a preset time period, and judging whether the average correction error exceeds a dynamic adjustment threshold; If it exceeds, adjusting the coefficient of the non-linear relationship expression in the compensation curve generation function according to the distribution characteristics of the average correction error; Reapplying the adjusted compensation curve generation function to the subsequent processing of the temperature drift characteristic to realize the adaptive update of the non-linear compensation curve.

5. The method according to claim 1, wherein The obtaining the sensor aging factor preset in the intelligent monitor includes: Obtain the historical working duration of the intelligent monitor and the sensor aging weight allocation table; wherein, the sensor aging weight allocation table includes the aging influence weights corresponding to different working duration intervals; Match the corresponding aging influence weight from the sensor aging weight allocation table according to the historical working duration, and perform weighted processing on the reference aging factor based on the aging influence weight to generate a dynamically updated sensor aging factor.

6. The method according to claim 5, wherein The generation process of the sensor aging weight allocation table includes: Obtain the historical data of multiple reference intelligent monitors during the aging test stage, and the historical data includes the sensor performance decay indexes at different working durations; According to the correlation between the sensor performance decay index and the temperature drift correction error, divide the influence levels corresponding to different working duration intervals; Based on the influence levels, allocate corresponding aging influence weights to each working duration interval to generate the sensor aging weight allocation table.

7. The method according to claim 1, characterized in that The adoption of the adaptive compensation parameter optimization algorithm to perform iterative optimization processing on the compensation parameters of the non-linear compensation curve includes: Obtain the temperature drift correction error in the current iteration cycle, and compare the temperature drift correction error with a preset error threshold; If the temperature drift correction error is greater than the preset error threshold, adjust the parameter values in the compensation parameter set according to the error gradient direction to generate an updated compensation parameter set; Re-input the updated compensation parameter set into the non-linear compensation curve for parameter verification until the temperature drift correction error is less than or equal to the preset error threshold.

8. The method according to claim 1, characterized in that, During the iterative optimization process, the integration of the sensor aging factor to dynamically correct the compensation parameters includes: Divide the preset sensor aging factor value intervals according to the numerical size of the sensor aging factor, and configure corresponding compensation parameter correction weights for each sensor aging factor value interval; Extract the offset compensation parameter and the temperature drift slope compensation parameter in the current iteration cycle from the compensation parameter set, and decompose the offset compensation parameter and the temperature drift slope compensation parameter into independent parameter components according to the parameter type; wherein, the independent parameter component is a dimensionless standardized parameter component obtained after the dimension reduction process; Match the corresponding compensation parameter correction weight according to the sensor aging factor value interval where the sensor aging factor is located, and allocate the compensation parameter correction weight to each independent parameter component; Based on the parameter change direction of each independent parameter component, perform directional weighted operation on the allocated compensation parameter correction weight to generate an aging correction component corresponding to each independent parameter component; Superimpose the aging correction component onto the corresponding independent parameter component to generate a corrected offset compensation parameter component and a corrected temperature drift slope compensation parameter component; Re-combine the corrected offset compensation parameter component and the corrected temperature drift slope compensation parameter component according to the parameter type to generate a dynamically corrected compensation parameter set; Input the dynamically corrected set of compensation parameters into the non-linear compensation curve for parameter verification, and obtain the verified temperature drift correction error; According to the error distribution characteristics of the verified temperature drift correction error, inversely adjust the compensation parameter correction weights corresponding to the numerical range of the sensor aging factor, and generate an updated compensation parameter correction weight table; Based on the updated compensation parameter correction weight table, perform dynamic correction weight allocation on the independent parameter components in the next iteration cycle to generate an optimized set of compensation parameters.

9. The method according to claim 1, characterized in that, The real-time correction process of the original measurement data of the intelligent monitor according to the optimized set of compensation parameters, and output the target measurement result after suppressing the temperature drift noise, includes: Divide the original measurement data into multiple time-synchronized measurement data segments according to the acquisition timestamp, and the time range of each measurement data segment corresponds one-to-one with the time range of the environmental temperature data segment; According to the time range of the environmental temperature data segment, match the corresponding set of compensation parameters from the optimized set of compensation parameters, and assign the set of compensation parameters to the time-synchronized measurement data segment; Based on the offset correction coefficient and the temperature drift slope in the set of compensation parameters, perform a linear compensation operation on each measurement value in the time-synchronized measurement data segment to generate a compensated measurement data sequence; Perform a sliding window mean filtering process on the compensated measurement data sequence, and calculate the average value of the measurement data in each window according to the preset window length and overlapping step size to generate a filtered measurement data sequence; Merge the filtered measurement data sequence in the order of timestamps, and remove the measurement data corresponding to the repeated timestamps to generate the target measurement result after suppressing the temperature drift noise.

10. An intelligent monitor temperature drift correction system, characterized in that, It includes a processor, a memory and a bus connected to the processor; wherein, the processor and the memory complete communication with each other through the bus; the processor is used to call the program instructions in the memory to execute the temperature drift correction method of the intelligent monitor based on the temperature compensation algorithm according to any one of claims 1-9.

Images