Sensing equipment operation control system and method for health detection

Abstract

The invention discloses a sensing equipment operation control system and method for health detection, and relates to the technical field of blood glucose monitoring. Historical data containing drift calibration information and reference data of equipment of the same specification of the equipment are extracted by obtaining a unique identifier of the sensing equipment and interacting with a server database; an individual and group sensing equipment drift set is constructed, and the problem that data lacks unique traceability and reference is solved; a to-be-predicted sequence is generated based on real-time calibration information, individual and reference drift prediction curves are obtained through a long-short-term memory network model, a comprehensive drift curve is formed through weighted fusion, and the problems of inaccurate prediction and precision degradation are solved; the method comprises the following steps: setting a weight adjustment period, calculating an error, correcting a weight in combination with a historical standard deviation, and calculating the service life of the next equipment based on the corrected weight and a drift limit value, thereby solving the problems of fixed weight and difficult control of service time in the prior art, realizing dynamic adjustment of the weight and accurate analysis of the service life, and ensuring the reliability of blood glucose monitoring.

Description

Technical Field

[0001] This invention relates to the field of blood glucose monitoring technology, specifically to a sensor device operation control system and method for health monitoring. Background Technology

[0002] Blood glucose monitoring is a core tool for precise blood glucose management in diabetic patients and those with abnormal blood glucose levels. Its importance lies in: real-time capture of blood glucose fluctuations, enabling timely warnings and avoidance of acute risks such as coma and convulsions caused by hypoglycemia, and ketoacidosis and hyperosmolar state caused by hyperglycemia, thus preventing immediate damage to organs caused by transient blood glucose abnormalities; long-term continuous blood glucose monitoring can dynamically reflect blood glucose control trends, providing objective data support for doctors to assess the condition, formulate or adjust insulin injection dosages, oral hypoglycemic drug regimens, and optimize dietary structure and exercise intensity, effectively preventing microvascular complications such as retinopathy, nephropathy, and neuropathy, as well as macrovascular complications such as coronary heart disease and stroke caused by long-term blood glucose dysregulation. It plays an irreplaceable key role in maintaining stable blood glucose levels, improving quality of life, reducing disease-related disability and mortality rates, and alleviating the medical burden.

[0003] Glucose oxidase (GOD) electrochemical sensors are the mainstream devices in the field of blood glucose detection. They rely on enzymatic reactions and then convert electrochemical signals to achieve real-time blood glucose detection, offering advantages such as fast response and low detection limit. However, glucose oxidase activity is easily affected by factors such as the subcutaneous microenvironment, the stability of the enzyme immobilization layer, and the duration of use, making it difficult to control the service life of the sensor. Furthermore, the sensor can only achieve a limited number of enzyme activity drift calibrations throughout its entire lifespan, resulting in insufficient calibration benchmarks to correct errors in the monitoring data. This leads to data drift, deterioration in detection accuracy, and affects the reliability of blood glucose management. Therefore, there is an urgent need for a control system and method for the operation of sensing devices used in health monitoring. Summary of the Invention

[0004] The purpose of this invention is to provide a control system and method for the operation of a sensing device for health monitoring, so as to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for controlling the operation of a sensing device for health monitoring, the method comprising the following steps: Step S1: Obtain the unique identifier of the sensing device, and use the unique identifier to obtain historical monitoring data to construct an individual sensing device drift analysis set, and obtain historical monitoring data of other sensing devices of the same specifications and models to construct a group sensing device drift reference set; Step S1-1: Obtain the unique identifier of the sensing device through the built-in non-volatile memory of the sensing device. The unique identifier is an unmodifiable serial number written to the sensing device at the factory. The sensing device interacts with the server database through the built-in communication component. Step S1-2: Read multiple sets of historical monitoring data from the server database according to the unique identifier of the sensing device. The historical monitoring data includes the timestamp of each blood glucose measurement, the blood glucose measurement value, and drift calibration information. The drift calibration information represents the calibration timestamp corresponding to the glucose oxidase electrochemical sensor when it is drift calibrated, the reference blood glucose value used, and the drift correction amount determined by calibration. The drift correction amount represents the deviation between the reference blood glucose value obtained at the same time and the actual blood glucose value measured by the sensor at a specific calibration time. A set of historical monitoring data specifically includes all blood glucose measurement records of the glucose oxidase electrochemical sensor from the start of its service to the end of a single sensor replacement. Step S1-3: Filter out historical monitoring data of other sensing devices that are consistent with the specifications and models of the sensing device from the server database, and record them as reference historical monitoring data. Steps S1-4: Select the unique identifier of the sensing device as the key and the historical monitoring data of the sensing device as the value to construct an individual sensing device drift analysis set; select the unique identifier of other sensing devices as the key and the corresponding reference historical monitoring data of other sensing devices as the value to construct a group sensing device drift reference set. The storage form of the group sensing device drift reference set is obtained by combining multiple individual sensing device drift analysis sets. By obtaining the unique identifier of the sensing device and establishing data interaction with the server database, historical monitoring data containing drift calibration information of the device and reference historical monitoring data of devices of the same specification and model are extracted. Individual sensor device drift analysis sets and group sensor device drift reference sets are constructed respectively, providing a unique and accurate data foundation with the same specification for subsequent sensor device drift analysis.

[0006] Step S2: Analyze the drift analysis set of individual sensors and the drift reference set of group sensors, extract the actual monitored blood glucose values, timestamps and drift calibration information, and construct drift curves for individual sensors and group sensors respectively. Step S2-1: Extract all drift calibration information of the sensor from the drift analysis set of individual sensor devices, and separate the calibration timestamp and drift correction amount corresponding to each drift calibration; extract the start service timestamp of the sensor device, calculate the time difference between each calibration timestamp and the start service timestamp, and use the time difference as the service progress parameter of the individual sensor device. Step S2-2: Select the service progress parameter of the individual sensor device as the horizontal axis to establish the x-axis, and select the corresponding drift correction amount as the vertical axis to establish the y-axis. Map the calibration data of the sensor device according to the x-axis and y-axis. When there are multiple drift correction amounts corresponding to the same service progress parameter, calculate the arithmetic mean of the group of drift correction amounts, and use it as the final drift correction amount corresponding to the service progress parameter. Step S2-3: Sort the service progress parameters of the sensor after processing by numerical value from smallest to largest, arrange the corresponding final drift correction amount in the same order, connect each data point in sequence, and construct the drift curve of individual sensor. Step S2-4: Extract the drift calibration information and start service timestamp of each sensor from the group sensor drift reference set, and repeat the processing steps S2-1 to S2-2 to obtain the calibration data of each sensor with the service progress parameter as the x-axis and the final drift correction amount as the y-axis. Step S2-5: Aggregate the calibration data of all sensors in the group sensor drift reference set according to the service progress parameter. When there are multiple final drift corrections corresponding to the same service progress parameter, calculate the arithmetic mean of the final drift corrections in the group and use it as the drift correction for other sensors corresponding to the same service progress parameter. Sort the service progress parameters of other sensors in ascending order of value, arrange their corresponding drift corrections in the same order, and connect the data points in sequence to construct the group sensor drift curve. Based on the drift analysis set of individual sensors and the drift reference set of group sensors, drift calibration information is first extracted from the individual set. The time difference between the calibration timestamp and the start of service timetamp is calculated as the service progress parameter. After coordinate mapping, averaging the drift correction under the same parameters, and sorting and connecting the data points, an individual sensor drift curve is constructed. Then, information of each device is extracted from the group set, and the individual data processing process is repeated. Data is aggregated according to the service progress parameter, and the average value is taken under the same parameters. The data points are sorted and connected to construct a group sensor drift curve, realizing the standardized visualization of drift data and providing accurate curve basis for subsequent sensor drift prediction.

[0007] Step S3: Obtain drift calibration information of the current service stage of the sensing device, analyze it in combination with the drift curve of individual sensing devices and the drift curve of group sensing devices, and predict the drift prediction curve of individual sensing devices and the drift prediction curve of reference sensing devices. Step S3-1: Obtain real-time drift calibration information from the current service stage of the sensing device; extract the corresponding calibration timestamp and drift correction amount based on the real-time drift calibration information; calculate the time difference between the calibration timestamp and the start time of the sensing device's service to obtain the current service progress parameters of the sensing device. Step S3-2: Map the current service progress parameters and corresponding drift corrections to the coordinate system of the drift curve of individual sensors to form new individual data points. Select the new individual data points as the reference endpoints to form the individual calibration data sequence to be predicted. Map the current service progress parameters and corresponding drift corrections to the coordinate system of the drift curve of the group sensors to form new group data points. Select the new group data points as the reference endpoints to form the group calibration data sequence to be predicted. Step S3-3: Based on the individual calibration data sequence to be predicted, the first long short-term memory network model is trained through the drift curve of the individual sensing device to obtain the mapping relationship between the service progress parameters and the corresponding drift correction in the individual calibration data sequence to be predicted. The input of the first long short-term memory network model is the individual calibration data sequence to be predicted, and the output is the predicted value of the drift correction of the sensing device under the subsequent service progress parameters. The predicted values ​​are connected in ascending order of the service progress parameters to obtain the drift prediction curve of the individual sensing device. Step S3-4: Based on the calibration data sequence of the group to be predicted, the second long short-term memory network model is trained through the drift curve of the group sensor device to obtain the mapping relationship between the service progress parameter and the corresponding drift correction in the calibration data sequence of the group to be predicted. The input of the second long short-term memory network model is the calibration data sequence of the group to be predicted, and the output is the predicted drift correction value of other sensor devices of the same model under the subsequent service progress parameter. The predicted values ​​are connected in ascending order of service progress parameter to obtain the drift prediction curve of the reference sensor device. By acquiring real-time drift calibration information of the current service stage of the sensing device, calculating the current service progress parameters and mapping them to the drift curves of individual and group sensing devices to form a data sequence to be predicted, and then training the individual and group calibration data sequences to be predicted using the first and second long short-term memory network models respectively, outputting the predicted drift correction values ​​under the subsequent service progress parameters and connecting the points, the drift prediction curve of the individual sensing device and the drift prediction curve of the reference sensing device are obtained, providing accurate individual and group drift prediction basis for the subsequent generation of comprehensive drift curve.

[0008] Step S4: Based on the drift calibration information processed by time alignment during the current service stage of the sensing device, the weight values ​​of the drift prediction curve of the individual sensing device and the weight values ​​of the drift prediction curve of the reference sensing device are obtained. The comprehensive drift curve of the sensing device is obtained through weighted fusion calculation. Step S4-1: Obtain the drift calibration information of the sensing device in real time, extract the drift correction amount and the corresponding calibration timestamp, and record them as the real-time drift correction amount and the real-time calibration timestamp respectively. Find the predicted value of the drift correction amount corresponding to the real-time calibration timestamp in the drift prediction curve of the individual sensing device, and calculate the difference between the real-time drift correction amount and the predicted value of the drift correction amount, and record it as the individual drift correction amount difference. Step S4-2: Simultaneously, find the predicted drift correction value corresponding to the real-time calibration timestamp in the drift prediction curve of the reference sensor, calculate the difference between the real-time drift correction and the predicted drift correction value, and record it as the group drift correction difference. Step S4-3: Perform a summation operation on the individual drift correction difference and the group drift correction difference to obtain the total drift correction difference; divide the individual drift correction difference by the total drift correction difference to obtain the weight value of the individual sensor drift prediction curve, denoted as the individual weight value S1; divide the group drift correction difference by the total drift correction difference to obtain the weight value of the reference sensor drift prediction curve, denoted as the reference weight value C1; combine the weighted fusion operation to generate the comprehensive drift curve of the sensor. By acquiring real-time drift calibration information of sensing devices, the real-time drift correction amount and the corresponding real-time calibration timestamp are extracted. The drift correction prediction value corresponding to the real-time calibration timestamp is found in the drift prediction curve of individual sensing devices and the drift prediction curve of reference sensing devices, respectively. The difference between the real-time drift correction amount and the individual drift correction amount, as well as the difference between the real-time drift correction amount and the group drift correction amount, are calculated. The total drift correction difference is obtained by summing the two types of differences. Then, the weight values ​​of the individual sensor drift prediction curve and the reference sensor drift prediction curve are obtained by dividing the two types of differences by the total difference, respectively. The comprehensive drift curve of the sensing devices is generated by weighted fusion calculation, which provides accurate comprehensive drift basis for subsequent sensor monitoring and life analysis.

[0009] Step S5: Set the weight value adjustment cycle to monitor and analyze the comprehensive drift curve of the sensing device, obtain the drift calibration information of the sensing device in real time, and calculate multiple absolute errors. Based on the analysis of multiple absolute errors, correct the weight value of the drift curve of the individual sensing device and the weight value of the drift curve of the reference sensing device. Analyze and calculate the service life of the sensing device based on the corrected weight value. Step S5-1: Set the weight value adjustment cycle. The weight value adjustment cycle is the cycle after N complete service cycles of the same specification and model of sensor devices are completed. The weight value is adjusted again after the weight adjustment cycle is completed. Based on the weight adjustment cycle, the comprehensive drift curve is analyzed to obtain the service life of the sensor device at the end of the weight value adjustment cycle. It is recorded as the verification sensor device. The drift correction amount is obtained by the verification sensor device at the end of its service life. The drift correction amount is recorded as the final drift correction amount. The time corresponding to the final drift correction amount is recorded as the final timestamp. Step S5-2: Extract the predicted drift correction amount corresponding to the final timestamp in the drift curve of the individual sensor device, and calculate the absolute error between the final drift correction amount and the predicted drift correction amount, denoted as the individual error H. single ; Step S5-3: Extract the predicted drift correction amount corresponding to the final timestamp in the drift curve of the reference sensing device, calculate the error between the final drift correction amount and the predicted drift value, and denot it as the reference error H. crowd ; Step S5-4: Based on the individual weight value S1, the reference weight value C1, and the individual error H... single and reference error H crowd The weight values ​​of the drift curve of the individual sensor and the drift curve of the reference sensor are corrected. The specific process is as follows: Step S5-4-1: Calculate the standard deviation f of multiple historical final drift corrections in the drift curve of an individual sensor device, and the standard deviation F of multiple historical final drift corrections in the drift curve of a reference sensor device; Step S5-4-2, the formula for calculating the adjustment coefficient of the individual weight value is as follows: k s =exp(-H single ÷f); In the formula, k s The adjustment coefficient is expressed as the individual weight value; Step S5-4-3, the formula for calculating the adjustment coefficient of the reference weight value is as follows: k H =exp(-H crowd ÷F); In the formula, k H This is expressed as an adjustment coefficient based on the reference weight value; Step S5-4-4, the formula for correcting individual weight values ​​is as follows: S 1，new =S1×k s ÷(S1×k s +C1×k H ); In the formula, S 1，new Represented as the revised individual weight values; using 1-S 1，new The new reference weight value C is obtained after correction. 1，new ; Step S5-5: Based on the corrected individual weight value S 1，new and the corrected reference weight value C 1，new After the next sensor of the same specification is drift-calibrated, a comprehensive drift curve is calculated. Based on the drift limit value specified for the sensor, which is defined as the maximum drift threshold that the sensor is allowed to have after calibration throughout its entire service life, the time when the next sensor of the same specification reaches the drift limit value is obtained, which is the service life of the sensor. By setting a weight adjustment cycle for every N complete service cycles of the same model of sensor, the service life of the calibrated sensor is obtained based on the comprehensive drift curve of the cycle analysis. The final drift correction amount and corresponding final timestamp at the end of its service are obtained. The individual error and reference error of the correction amount compared with the predicted value of the corresponding timestamp in the drift curve of the individual and reference sensor are calculated. The adjustment coefficient is calculated by combining the standard deviation of the historical final drift correction amount of the individual and reference curves, and the weight values ​​of the individual and reference curves are corrected. Then, based on the corrected weights, the comprehensive drift curve of the next sensor of the same model after drift calibration is calculated. Combined with the drift limit value of the maximum allowable drift amount, the time when the device reaches the limit value, i.e., the service life, is obtained, realizing dynamic adjustment of weights and accurate analysis of service life.

[0010] Furthermore, a sensor device operation control system for health monitoring includes a drift data set construction module, a drift curve plotting module, a drift curve prediction module, a comprehensive drift curve generation module, and a service life analysis module. The drift data set construction module is used to construct an individual sensor drift analysis set and a group sensor drift reference set; the drift curve plotting module is used to construct individual sensor drift curves and group sensor drift curves; the drift curve prediction module is used to predict the drift curves of individual sensors and reference sensors; the comprehensive drift curve generation module is used to generate a comprehensive drift curve of the sensors; and the service life analysis module is used to analyze and calculate the service life of the sensors. The output of the drift data set construction module is electrically connected to the input of the drift curve plotting module; the output of the drift curve plotting module is electrically connected to the input of the drift curve prediction module; the output of the drift curve prediction module is electrically connected to the input of the comprehensive drift curve generation module; and the output of the comprehensive drift curve generation module is electrically connected to the input of the service life analysis module. The drift data set construction module includes an identification data acquisition unit and a set construction unit; the identification data acquisition unit is used to acquire the unique identifier of the sensing device, the historical monitoring data of the individual sensing device, and the historical monitoring data of other sensing devices of the same specification and model; the set construction unit is used to construct an individual sensing device drift analysis set and a group sensing device drift reference set, respectively, using the unique identifier of the sensing device as the key and the corresponding historical monitoring data as the value. The drift curve plotting module includes a calibration information extraction unit and a curve plotting unit. The calibration information extraction unit is used to extract drift calibration information from the drift analysis set of individual sensing devices and the drift reference set of group sensing devices, calculate service progress parameters, and process drift correction amounts. The curve plotting unit is used to sort the processed service progress parameters and drift correction amounts, and connect the data points sequentially to construct drift curves for individual sensing devices and drift curves for group sensing devices. The drift curve prediction module includes a current calibration acquisition unit and a curve prediction unit. The current calibration acquisition unit is used to acquire drift calibration information of the current service stage of the sensing device, calculate the current service progress parameters, and form a calibration data sequence of the individual to be predicted and a calibration data sequence of the group to be predicted. The curve prediction unit is used to train the calibration data sequence of the individual to be predicted through a first long short-term memory network model and train the calibration data sequence of the group to be predicted through a second long short-term memory network model, and generate the drift curve of the individual sensing device and the drift curve of the reference sensing device, respectively. The comprehensive drift curve generation module includes a weight calculation unit and a comprehensive curve generation unit; the weight calculation unit is used to calculate the weight value S1 of the drift curve of the individual sensing device and the weight value C1 of the drift curve of the reference sensing device; the comprehensive curve generation unit is used to perform a weighted fusion operation on the drift curve of the individual sensing device and the drift curve of the reference sensing device using the weight value S1 and the weight value C1 to generate a comprehensive drift curve. The service life analysis module includes a weight adjustment unit and a service life determination unit; the weight adjustment unit is used to set the weight value adjustment period and calculate the individual error H. single With reference error H crowd The individual weight value S is obtained after correction. 1，new Compared with the reference weight value C 1，new The lifespan determination unit is used to determine the lifespan based on the individual weight value S. 1，new Compared with the reference weight value C 1，new Calculate the overall drift curve of the next sensor of the same specification and model, and combine it with the drift limit value to obtain the service life of the sensor.

[0011] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention obtains an unmodifiable unique identifier from the built-in non-volatile memory of the sensing device, interacts with the server database through communication components, and extracts historical monitoring data containing drift calibration information of the device and reference historical monitoring data of devices of the same specification and model. It then constructs an individual sensor device drift analysis set and a group sensor device drift reference set, respectively. This solves the problem of lack of unique traceability and reference of the same specification in the prior art, and provides a unique and comprehensive accurate data foundation for subsequent drift analysis, avoiding analysis errors caused by data confusion.

[0012] 2. This invention obtains real-time drift calibration information of the current service stage of the sensing device, calculates the current service progress parameters and maps them to individual and group drift curves to form a sequence to be predicted. The first and second long short-term memory network models are used to train and generate drift prediction curves for individual and reference sensing devices, respectively. Then, the error is calculated and weighted to generate a comprehensive drift curve. This solves the problems of inaccurate drift prediction and deterioration of detection accuracy in the prior art, and provides accurate comprehensive drift basis for monitoring and life analysis.

[0013] 3. This invention sets a weight adjustment period for every N complete service cycles of devices of the same specification and model, obtains the final drift correction amount and timestamp at the end of the service life of the verification device, calculates the error and corrects the weight by combining the historical standard deviation, and calculates the service life of the next device based on the corrected weight and drift limit value. This solves the problems of fixed weight and difficult control of service time in the prior art, realizes dynamic adjustment of weight and accurate analysis of service life, and ensures the reliability of blood glucose monitoring. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the operation control method of a sensing device for health monitoring according to the present invention. Figure 2 This is a schematic diagram of the operation control system of a sensing device for health detection according to the present invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Example 1: As Figure 1 As shown, the present invention provides a technical solution: a method for controlling the operation of a sensing device for health monitoring. The method includes the following steps: Step S1: Obtain the unique identifier of the sensing device, and use the unique identifier to obtain historical monitoring data to construct an individual sensing device drift analysis set, and obtain historical monitoring data of other sensing devices of the same specifications and models to construct a group sensing device drift reference set; Step S1-1: Obtain the unique identifier of the sensing device through the built-in non-volatile memory of the sensing device. The unique identifier is an unmodifiable serial number written to the sensing device at the factory. The sensing device interacts with the server database through the built-in communication component. Step S1-2: Read multiple sets of historical monitoring data from the server database according to the unique identifier of the sensing device. The historical monitoring data includes the timestamp of each blood glucose measurement, the blood glucose measurement value, and drift calibration information. The drift calibration information represents the calibration timestamp corresponding to the glucose oxidase electrochemical sensor when it is drift calibrated, the reference blood glucose value used, and the drift correction amount determined by calibration. The drift correction amount represents the deviation between the reference blood glucose value obtained at the same time and the actual blood glucose value measured by the sensor at a specific calibration time. A set of historical monitoring data specifically includes all blood glucose measurement records of the glucose oxidase electrochemical sensor from the start of its service to the end of a single sensor replacement. Step S1-3: Filter out historical monitoring data of other sensing devices that are consistent with the specifications and models of the sensing device from the server database, and record them as reference historical monitoring data. Steps S1-4: Select the unique identifier of the sensing device as the key and the historical monitoring data of the sensing device as the value to construct an individual sensing device drift analysis set; select the unique identifier of other sensing devices as the key and the corresponding reference historical monitoring data of other sensing devices as the value to construct a group sensing device drift reference set. The storage form of the group sensing device drift reference set is obtained by combining multiple individual sensing device drift analysis sets. In practical implementation, the unique identifier of the sensing device can be obtained by reading its built-in non-volatile memory. This identifier, as an unmodifiable serial number, is the core basis for device data traceability. The communication component needs to stably achieve bidirectional interaction with the server database to ensure the complete extraction of historical monitoring data. When screening historical data of devices of the same specifications and models, the device model parameters must be strictly matched to avoid affecting the effectiveness of the group reference set due to specification differences. Using the unique identifier as the key when constructing the set can ensure the accuracy of data retrieval.

[0017] Step S2: Analyze the drift analysis set of individual sensors and the drift reference set of group sensors, extract the actual monitored blood glucose values, timestamps and drift calibration information, and construct drift curves for individual sensors and group sensors respectively. Step S2-1: Extract all drift calibration information of the sensor from the drift analysis set of individual sensor devices, and separate the calibration timestamp and drift correction amount corresponding to each drift calibration; extract the start service timestamp of the sensor device, calculate the time difference between each calibration timestamp and the start service timestamp, and use the time difference as the service progress parameter of the individual sensor device. Step S2-2: Select the service progress parameter of the individual sensor device as the horizontal axis to establish the x-axis, and select the corresponding drift correction amount as the vertical axis to establish the y-axis. Map the calibration data of the sensor device according to the x-axis and y-axis. When there are multiple drift correction amounts corresponding to the same service progress parameter, calculate the arithmetic mean of the group of drift correction amounts, and use it as the final drift correction amount corresponding to the service progress parameter. Step S2-3: Sort the service progress parameters of the sensor after processing by numerical value from smallest to largest, arrange the corresponding final drift correction amount in the same order, connect each data point in sequence, and construct the drift curve of individual sensor. Step S2-4: Extract the drift calibration information and start service timestamp of each sensor from the group sensor drift reference set, and repeat the processing steps S2-1 to S2-2 to obtain the calibration data of each sensor with the service progress parameter as the x-axis and the final drift correction amount as the y-axis. Step S2-5: Aggregate the calibration data of all sensors in the group sensor drift reference set according to the service progress parameter. When there are multiple final drift corrections corresponding to the same service progress parameter, calculate the arithmetic mean of the final drift corrections in the group and use it as the drift correction for other sensors corresponding to the same service progress parameter. Sort the service progress parameters of other sensors in ascending order of value, arrange their corresponding drift corrections in the same order, and connect the data points in sequence to construct the group sensor drift curve. In practical implementation, constructing an individual sensor drift curve requires first clarifying the calculation logic of the service progress parameter, namely the difference between the calibration timestamp and the start of service timestamp, to unify the time dimension; taking the average of multiple drift corrections corresponding to the same service progress parameter can weaken the impact of random errors on the curve; when sorting and connecting points to form a curve, the continuity of data points must be maintained to ensure that the curve can truly reflect the individual drift trend; the construction of a group curve requires aggregating data from devices of the same specification, extracting common drift characteristics of the group through mean calculation, and ensuring comparability with individual curves.

[0018] Step S3: Obtain drift calibration information of the current service stage of the sensing device, analyze it in combination with the drift curve of individual sensing devices and the drift curve of group sensing devices, and predict the drift prediction curve of individual sensing devices and the drift prediction curve of reference sensing devices. Step S3-1: Obtain real-time drift calibration information from the current service stage of the sensing device; extract the corresponding calibration timestamp and drift correction amount based on the real-time drift calibration information; calculate the time difference between the calibration timestamp and the start time of the sensing device's service to obtain the current service progress parameters of the sensing device. Step S3-2: Map the current service progress parameters and corresponding drift corrections to the coordinate system of the drift curve of individual sensors to form new individual data points. Select the new individual data points as the reference endpoints to form the individual calibration data sequence to be predicted. Map the current service progress parameters and corresponding drift corrections to the coordinate system of the drift curve of the group sensors to form new group data points. Select the new group data points as the reference endpoints to form the group calibration data sequence to be predicted. Step S3-3: Based on the individual calibration data sequence to be predicted, the first long short-term memory network model is trained through the drift curve of the individual sensing device to obtain the mapping relationship between the service progress parameters and the corresponding drift correction in the individual calibration data sequence to be predicted. The input of the first long short-term memory network model is the individual calibration data sequence to be predicted, and the output is the predicted value of the drift correction of the sensing device under the subsequent service progress parameters. The predicted values ​​are connected in ascending order of the service progress parameters to obtain the drift prediction curve of the individual sensing device. Step S3-4: Based on the calibration data sequence of the group to be predicted, the second long short-term memory network model is trained through the drift curve of the group sensor device to obtain the mapping relationship between the service progress parameter and the corresponding drift correction in the calibration data sequence of the group to be predicted. The input of the second long short-term memory network model is the calibration data sequence of the group to be predicted, and the output is the predicted drift correction value of other sensor devices of the same model under the subsequent service progress parameter. The predicted values ​​are connected in ascending order of service progress parameter to obtain the drift prediction curve of the reference sensor device. In practical implementation, obtaining real-time drift calibration information for the current service stage requires relying on the equipment's real-time monitoring function to ensure the synchronization of calibration timestamps and drift correction amounts. Mapping the data to an existing curve coordinate system to form a sequence to be predicted can effectively connect new data with historical trends. When training a long short-term memory network model, the mapping relationship between service progress parameters and drift correction amounts should be the core. The matching of model input and output directly affects the reliability of the prediction curve. Individual and group models need to be trained independently to retain their respective features.

[0019] Step S4: Based on the drift calibration information processed by time alignment during the current service stage of the sensing device, the weight values ​​of the drift prediction curve of the individual sensing device and the weight values ​​of the drift prediction curve of the reference sensing device are obtained. The comprehensive drift curve of the sensing device is obtained through weighted fusion calculation. Step S4-1: Obtain the drift calibration information of the sensing device in real time, extract the drift correction amount and the corresponding calibration timestamp, and record them as the real-time drift correction amount and the real-time calibration timestamp respectively. Find the predicted value of the drift correction amount corresponding to the real-time calibration timestamp in the drift prediction curve of the individual sensing device, and calculate the difference between the real-time drift correction amount and the predicted value of the drift correction amount, and record it as the individual drift correction amount difference. Step S4-2: Simultaneously, find the predicted drift correction value corresponding to the real-time calibration timestamp in the drift prediction curve of the reference sensor, calculate the difference between the real-time drift correction and the predicted drift correction value, and record it as the group drift correction difference. Step S4-3: Perform a summation operation on the individual drift correction difference and the group drift correction difference to obtain the total drift correction difference; divide the individual drift correction difference by the total drift correction difference to obtain the weight value of the individual sensor drift prediction curve, denoted as the individual weight value S1; divide the group drift correction difference by the total drift correction difference to obtain the weight value of the reference sensor drift prediction curve, denoted as the reference weight value C1; combine the weighted fusion operation to generate the comprehensive drift curve of the sensor. In practical implementation, the extraction of real-time drift calibration information must be strictly bound to timestamps to ensure that the predicted values ​​found in the individual and group drift prediction curves correspond to the same time node. When calculating the difference between the individual and group drift corrections, attention must be paid to the positive or negative attribute of the difference, as it reflects the direction of deviation between the predicted and actual values. Weight calculation is based on the total difference, which makes the influence ratio of individuals and reference curves more reasonable. When generating a comprehensive curve through weighted fusion, it is necessary to ensure that the sum of the weight values ​​is 1 to guarantee the objectivity of the curve. Step S5: Set the weight value adjustment cycle to monitor and analyze the comprehensive drift curve of the sensing device, obtain the drift calibration information of the sensing device in real time, and calculate multiple absolute errors. Based on the analysis of multiple absolute errors, correct the weight value of the drift curve of the individual sensing device and the weight value of the drift curve of the reference sensing device. Analyze and calculate the service life of the sensing device based on the corrected weight value. Step S5-1: Set the weight value adjustment cycle. The weight value adjustment cycle is the cycle after N complete service cycles of the same specification and model of sensor devices are completed. The weight value is adjusted again after the weight adjustment cycle is completed. Based on the weight adjustment cycle, the comprehensive drift curve is analyzed to obtain the service life of the sensor device at the end of the weight value adjustment cycle. It is recorded as the verification sensor device. The drift correction amount is obtained by the verification sensor device at the end of its service life. The drift correction amount is recorded as the final drift correction amount. The time corresponding to the final drift correction amount is recorded as the final timestamp. Step S5-2: Extract the predicted drift correction amount corresponding to the final timestamp in the drift curve of the individual sensor device, and calculate the absolute error between the final drift correction amount and the predicted drift correction amount, denoted as the individual error H. single ; Step S5-3: Extract the predicted drift correction amount corresponding to the final timestamp in the drift curve of the reference sensing device, calculate the error between the final drift correction amount and the predicted drift value, and denot it as the reference error H. crowd ; Step S5-4: Based on the individual weight value S1, the reference weight value C1, and the individual error H...single and reference error H crowd The weight values ​​of the drift curve of the individual sensor and the drift curve of the reference sensor are corrected. The specific process is as follows: Step S5-4-1: Calculate the standard deviation f of multiple historical final drift corrections in the drift curve of an individual sensor device, and the standard deviation F of multiple historical final drift corrections in the drift curve of a reference sensor device; Step S5-4-2, the formula for calculating the adjustment coefficient of the individual weight value is as follows: k s =exp(-H single ÷f); In the formula, k s The adjustment coefficient is expressed as the individual weight value; Step S5-4-3, the formula for calculating the adjustment coefficient of the reference weight value is as follows: k H =exp(-H crowd ÷F); In the formula, k H This is expressed as an adjustment coefficient based on the reference weight value; Step S5-4-4, the formula for correcting individual weight values ​​is as follows: S 1，new =S1×k s ÷(S1×k s +C1×k H ); In the formula, S 1，new Represented as the revised individual weight values; using 1-S 1，new The new reference weight value C is obtained after correction. 1，new ; Step S5-5: Based on the corrected individual weight value S 1，new and the corrected reference weight value C 1，new After the next sensor of the same specification is drift-calibrated, a comprehensive drift curve is calculated. Based on the drift limit value specified for the sensor, which is defined as the maximum drift threshold that the sensor is allowed to have after calibration throughout its entire service life, the time when the next sensor of the same specification reaches the drift limit value is obtained, which is the service life of the sensor. In practical implementation, the weight adjustment cycle should be set in conjunction with the average service life of equipment of the same specifications, so that the cycle is neither too frequent to increase the computational load nor too long to cause weight lag; the final drift correction amount of the calibration sensor should be collected in real time at the end of its service life to ensure data timeliness; error calculation and weight correction should be correlated with historical standard deviation so that the adjustment coefficient can reasonably reflect the relative size of the error; when calculating the comprehensive curve based on the corrected weight and combining it with the drift limit value to evaluate the life, it is necessary to ensure that the limit value is consistent with the equipment performance standard to ensure the practicality of the life evaluation.

[0020] Example 2, as Figure 2 As shown, the present invention provides a sensor device operation control system for health detection, which includes a drift data set construction module, a drift curve plotting module, a drift curve prediction module, a comprehensive drift curve generation module, and a service life analysis module. The drift data set construction module is used to construct an individual sensor drift analysis set and a group sensor drift reference set; the drift curve plotting module is used to construct individual sensor drift curves and group sensor drift curves; the drift curve prediction module is used to predict the drift curves of individual sensors and reference sensors; the comprehensive drift curve generation module is used to generate a comprehensive drift curve of the sensors; and the service life analysis module is used to analyze and calculate the service life of the sensors. The output of the drift data set construction module is electrically connected to the input of the drift curve plotting module; the output of the drift curve plotting module is electrically connected to the input of the drift curve prediction module; the output of the drift curve prediction module is electrically connected to the input of the comprehensive drift curve generation module; and the output of the comprehensive drift curve generation module is electrically connected to the input of the service life analysis module. The drift data set construction module includes an identification data acquisition unit and a set construction unit; the identification data acquisition unit is used to acquire the unique identifier of the sensing device, the historical monitoring data of the individual sensing device, and the historical monitoring data of other sensing devices of the same specification and model; the set construction unit is used to construct an individual sensing device drift analysis set and a group sensing device drift reference set, respectively, using the unique identifier of the sensing device as the key and the corresponding historical monitoring data as the value. The drift curve plotting module includes a calibration information extraction unit and a curve plotting unit. The calibration information extraction unit is used to extract drift calibration information from the drift analysis set of individual sensing devices and the drift reference set of group sensing devices, calculate service progress parameters, and process drift correction amounts. The curve plotting unit is used to sort the processed service progress parameters and drift correction amounts, and connect the data points sequentially to construct drift curves for individual sensing devices and drift curves for group sensing devices. The drift curve prediction module includes a current calibration acquisition unit and a curve prediction unit. The current calibration acquisition unit is used to acquire drift calibration information of the current service stage of the sensing device, calculate the current service progress parameters, and form a calibration data sequence of the individual to be predicted and a calibration data sequence of the group to be predicted. The curve prediction unit is used to train the calibration data sequence of the individual to be predicted through a first long short-term memory network model and train the calibration data sequence of the group to be predicted through a second long short-term memory network model, and generate the drift curve of the individual sensing device and the drift curve of the reference sensing device, respectively. The comprehensive drift curve generation module includes a weight calculation unit and a comprehensive curve generation unit; the weight calculation unit is used to calculate the weight value S1 of the drift curve of the individual sensing device and the weight value C1 of the drift curve of the reference sensing device; the comprehensive curve generation unit is used to perform a weighted fusion operation on the drift curve of the individual sensing device and the drift curve of the reference sensing device using the weight value S1 and the weight value C1 to generate a comprehensive drift curve. The service life analysis module includes a weight adjustment unit and a service life determination unit; the weight adjustment unit is used to set the weight value adjustment period and calculate the individual error H. single With reference error H crowd The individual weight value S is obtained after correction. 1，new Compared with the reference weight value C 1，new The lifespan determination unit is used to determine the lifespan based on the individual weight value S. 1，new Compared with the reference weight value C 1，new Calculate the overall drift curve of the next sensor of the same specification and model, and combine it with the drift limit value to obtain the service life of the sensor.

[0021] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, it is intended that all variations falling within the meaning and scope of equivalents of the claims be included within the present invention.

Claims

1. A method for controlling the operation of a sensing device for health monitoring, characterized in that: The sensor operation control method includes the following steps: Step S1: Obtain the unique identifier of the sensing device, and use the unique identifier to obtain historical monitoring data to construct an individual sensing device drift analysis set, and obtain historical monitoring data of other sensing devices of the same specifications and models to construct a group sensing device drift reference set; Step S2: Analyze the drift analysis set of individual sensors and the drift reference set of group sensors, extract the actual monitored blood glucose values, timestamps and drift calibration information, and construct drift curves for individual sensors and group sensors respectively. Step S3: Obtain drift calibration information of the current service stage of the sensing device, analyze it in combination with the drift curve of individual sensing devices and the drift curve of group sensing devices, and predict the drift prediction curve of individual sensing devices and the drift prediction curve of reference sensing devices. Step S4: Based on the drift calibration information of the current service stage of the sensing device, the weight value of the drift prediction curve of the individual sensing device and the weight value of the drift prediction curve of the reference sensing device are obtained. The comprehensive drift curve of the sensing device is obtained through weighted fusion calculation. Step S5: Set the weight value adjustment cycle to monitor and analyze the comprehensive drift curve of the sensing device, obtain the drift calibration information of the sensing device in real time, and calculate multiple absolute errors. Based on the analysis of multiple absolute errors, correct the weight value of the drift curve of the individual sensing device and the weight value of the drift curve of the reference sensing device. Analyze and calculate the service life of the sensing device based on the corrected weight values.

2. The method for controlling the operation of a sensing device for health monitoring according to claim 1, characterized in that: The specific steps of step S1 are as follows: Step S1-1: Obtain the unique identifier of the sensing device through the built-in non-volatile memory of the sensing device. The unique identifier is an unmodifiable serial number written to the sensing device at the factory. The sensing device interacts with the server database through the built-in communication component. Step S1-2: Read multiple sets of historical monitoring data from the server database according to the unique identifier of the sensing device. The historical monitoring data includes the timestamp of each blood glucose measurement, the blood glucose measurement value, and drift calibration information. The drift calibration information represents the calibration timestamp corresponding to the glucose oxidase electrochemical sensor when it is drift calibrated, the reference blood glucose value used, and the drift correction amount determined by calibration. The drift correction amount represents the deviation between the reference blood glucose value obtained at the same time and the actual blood glucose value measured by the sensor at a specific calibration time. A set of historical monitoring data specifically includes all blood glucose measurement records of the glucose oxidase electrochemical sensor from the start of its service to the end of a single sensor replacement. Step S1-3: Filter out historical monitoring data of other sensing devices that are consistent with the specifications and models of the sensing device from the server database, and record them as reference historical monitoring data. Steps S1-4: Select the unique identifier of the sensing device as the key and the historical monitoring data of the sensing device as the value to construct an individual sensing device drift analysis set; select the unique identifier of other sensing devices as the key and the corresponding reference historical monitoring data of other sensing devices as the value to construct a group sensing device drift reference set. The storage form of the group sensing device drift reference set is obtained by combining multiple individual sensing device drift analysis sets.

3. The method for controlling the operation of a sensing device for health monitoring according to claim 2, characterized in that: The specific steps of step S2 are as follows: Step S2-1: Extract all drift calibration information of the sensor from the drift analysis set of individual sensor devices, and separate the calibration timestamp and drift correction amount corresponding to each drift calibration; extract the start service timestamp of the sensor device, calculate the time difference between each calibration timestamp and the start service timestamp, and use the time difference as the service progress parameter of the individual sensor device. Step S2-2: Select the service progress parameter of the individual sensor device as the horizontal axis to establish the x-axis, and select the corresponding drift correction amount as the vertical axis to establish the y-axis. Map the calibration data of the sensor device according to the x-axis and y-axis. When there are multiple drift correction amounts corresponding to the same service progress parameter, calculate the arithmetic mean of the group of drift correction amounts, and use it as the final drift correction amount corresponding to the service progress parameter. Step S2-3: Sort the processed service progress parameters of the sensing devices in ascending order of value, arrange the corresponding final drift corrections in the same order, connect the data points in sequence, and construct the drift curve of the individual sensing device.

4. The method for controlling the operation of a sensing device for health monitoring according to claim 3, characterized in that: Step S2 also includes: Step S2-4: Extract the drift calibration information and start service timestamp of each sensor from the group sensor drift reference set, and repeat the processing steps S2-1 to S2-2 to obtain the calibration data of each sensor with the service progress parameter as the x-axis and the final drift correction amount as the y-axis. Step S2-5: Aggregate the calibration data of all sensors in the group sensor drift reference set according to the service progress parameter. When there are multiple final drift corrections corresponding to the same service progress parameter, calculate the arithmetic mean of the final drift corrections in the group and use it as the drift correction for other sensors corresponding to the same service progress parameter. Sort the service progress parameters of other sensors in ascending order of value, arrange their corresponding drift corrections in the same order, and connect the data points in sequence to construct the group sensor drift curve.

5. The method for controlling the operation of a sensing device for health monitoring according to claim 4, characterized in that: The specific steps of step S3 are as follows: Step S3-1: Obtain real-time drift calibration information from the current service stage of the sensing device; extract the corresponding calibration timestamp and drift correction amount based on the real-time drift calibration information; calculate the time difference between the calibration timestamp and the start time of the sensing device's service to obtain the current service progress parameters of the sensing device. Step S3-2: Map the current service progress parameters and corresponding drift corrections to the coordinate system of the drift curve of individual sensors to form new individual data points. Select the new individual data points as the reference endpoints to form the individual calibration data sequence to be predicted. Map the current service progress parameters and corresponding drift corrections to the coordinate system of the drift curve of the group sensors to form new group data points. Select the new group data points as the reference endpoints to form the group calibration data sequence to be predicted. Step S3-3: Based on the individual calibration data sequence to be predicted, the first long short-term memory network model is trained through the drift curve of the individual sensing device to obtain the mapping relationship between the service progress parameters and the corresponding drift correction in the individual calibration data sequence to be predicted. The input of the first long short-term memory network model is the individual calibration data sequence to be predicted, and the output is the predicted value of the drift correction of the sensing device under the subsequent service progress parameters. The predicted values ​​are connected in ascending order of the service progress parameters to obtain the drift prediction curve of the individual sensing device. Step S3-4: Based on the calibration data sequence of the group to be predicted, the second long short-term memory network model is trained through the drift curve of the group sensor devices to obtain the mapping relationship between the service progress parameters and the corresponding drift correction in the calibration data sequence of the group to be predicted. The input of the second long short-term memory network model is the calibration data sequence of the group to be predicted, and the output is the predicted drift correction value of other sensor devices of the same model under the subsequent service progress parameters. The predicted values ​​are connected in ascending order of service progress parameters to obtain the drift prediction curve of the reference sensor device.

6. The method for controlling the operation of a sensing device for health monitoring according to claim 5, characterized in that: The specific steps of step S4 are as follows: Step S4-1: Obtain the drift calibration information of the sensing device in real time, extract the drift correction amount and the corresponding calibration timestamp, and record them as the real-time drift correction amount and the real-time calibration timestamp respectively. Find the predicted value of the drift correction amount corresponding to the real-time calibration timestamp in the drift prediction curve of the individual sensing device, and calculate the difference between the real-time drift correction amount and the predicted value of the drift correction amount, and record it as the individual drift correction amount difference. Step S4-2: Simultaneously, find the predicted drift correction value corresponding to the real-time calibration timestamp in the drift prediction curve of the reference sensor, calculate the difference between the real-time drift correction and the predicted drift correction value, and record it as the group drift correction difference. Step S4-3: Perform a summation operation on the individual drift correction difference and the group drift correction difference to obtain the total drift correction difference; divide the individual drift correction difference by the total drift correction difference to obtain the weight value of the individual sensor drift prediction curve, denoted as the individual weight value S1; divide the group drift correction difference by the total drift correction difference to obtain the weight value of the reference sensor drift prediction curve, denoted as the reference weight value C1; combine the weighted fusion operation to generate the comprehensive drift curve of the sensor.

7. The method for controlling the operation of a sensing device for health monitoring according to claim 6, characterized in that: The specific steps of step S5 are as follows: Step S5-1: Set the weight value adjustment cycle. The weight value adjustment cycle is the cycle after N complete service cycles of the same specification and model of sensor devices are completed. The weight value is adjusted again after the weight adjustment cycle is completed. Based on the weight adjustment cycle, the comprehensive drift curve is analyzed to obtain the service life of the sensor device at the end of the weight value adjustment cycle. It is recorded as the verification sensor device. The drift correction amount is obtained by the verification sensor device at the end of its service life. The drift correction amount is recorded as the final drift correction amount. The time corresponding to the final drift correction amount is recorded as the final timestamp. Step S5-2: Extract the predicted drift correction amount corresponding to the final timestamp in the drift curve of the individual sensor device, and calculate the absolute error between the final drift correction amount and the predicted drift correction amount, denoted as the individual error H. single ; Step S5-3: Extract the predicted drift correction amount corresponding to the final timestamp in the drift curve of the reference sensing device, calculate the error between the final drift correction amount and the predicted drift value, and denot it as the reference error H. crowd ; Step S5-4: Based on the individual weight value S1, the reference weight value C1, and the individual error H... single and reference error H crowd The weight values ​​of the drift curve of the individual sensor and the drift curve of the reference sensor are corrected. The specific process is as follows: Step S5-4-1: Calculate the standard deviation f of multiple historical final drift corrections in the drift curve of an individual sensor device, and the standard deviation F of multiple historical final drift corrections in the drift curve of a reference sensor device; Step S5-4-2, the formula for calculating the adjustment coefficient of the individual weight value is as follows: k s =exp(-H single ÷f); In the formula, k s The adjustment coefficient is expressed as the individual weight value; Step S5-4-3, the formula for calculating the adjustment coefficient of the reference weight value is as follows: k H =exp(-H crowd ÷F); In the formula, k H This is expressed as an adjustment coefficient based on the reference weight value; Step S5-4-4, the formula for correcting individual weight values ​​is as follows: S 1，new =S1×k s ÷(S1×k s +C1×k H ); In the formula, S 1，new Represented as the revised individual weight values; using 1-S 1，new The new reference weight value C is obtained after correction. 1，new ; Step S5-5: Based on the corrected individual weight value S 1，new and the corrected reference weight value C 1，new After the next sensor of the same specification is drift-calibrated, a comprehensive drift curve is calculated. Based on the drift limit value specified for the sensor, which is the threshold value that defines the maximum amount of drift that the sensor is allowed to have after calibration throughout its entire service life, the time when the next sensor of the same specification reaches the drift limit value is obtained, thus obtaining the service life of the sensor.

8. A sensor device operation control system for health monitoring, applied to the sensor device operation control method for health monitoring as described in any one of claims 1-7, characterized in that: The sensor device operation control system includes a drift data set construction module, a drift curve plotting module, a drift curve prediction module, a comprehensive drift curve generation module, and a service life analysis module. The drift data set construction module is used to construct an individual sensor drift analysis set and a group sensor drift reference set; the drift curve plotting module is used to construct individual sensor drift curves and group sensor drift curves; the drift curve prediction module is used to predict the drift curves of individual sensors and reference sensors; the comprehensive drift curve generation module is used to generate a comprehensive drift curve of the sensors; and the service life analysis module is used to analyze and calculate the service life of the sensors. The output of the drift data set construction module is electrically connected to the input of the drift curve plotting module; the output of the drift curve plotting module is electrically connected to the input of the drift curve prediction module; the output of the drift curve prediction module is electrically connected to the input of the comprehensive drift curve generation module; and the output of the comprehensive drift curve generation module is electrically connected to the input of the service life analysis module.

9. The operating control system for a sensing device for health monitoring according to claim 8, characterized in that: The drift data set construction module includes an identification data acquisition unit and a set construction unit; the identification data acquisition unit is used to acquire the unique identifier of the sensing device, the historical monitoring data of the individual sensing device, and the historical monitoring data of other sensing devices of the same specification and model; the set construction unit is used to construct an individual sensing device drift analysis set and a group sensing device drift reference set, respectively, using the unique identifier of the sensing device as the key and the corresponding historical monitoring data as the value. The drift curve plotting module includes a calibration information extraction unit and a curve plotting unit. The calibration information extraction unit is used to extract drift calibration information from the drift analysis set of individual sensing devices and the drift reference set of group sensing devices, calculate service progress parameters, and process drift correction amounts. The curve plotting unit is used to sort the processed service progress parameters and drift correction amounts, and connect the data points sequentially to construct drift curves for individual sensing devices and drift curves for group sensing devices. The drift curve prediction module includes a current calibration acquisition unit and a curve prediction unit. The current calibration acquisition unit is used to acquire drift calibration information of the current service stage of the sensing device, calculate the current service progress parameters, and form a calibration data sequence of the individual to be predicted and a calibration data sequence of the group to be predicted. The curve prediction unit is used to train the calibration data sequence of the individual to be predicted through a first long short-term memory network model and train the calibration data sequence of the group to be predicted through a second long short-term memory network model, and generate the drift curve of the individual sensing device and the drift curve of the reference sensing device, respectively.

10. The operating control system for a sensing device for health monitoring according to claim 8, characterized in that: The comprehensive drift curve generation module includes a weight calculation unit and a comprehensive curve generation unit; the weight calculation unit is used to calculate the weight value S1 of the drift curve of the individual sensing device and the weight value C1 of the drift curve of the reference sensing device; the comprehensive curve generation unit is used to perform a weighted fusion operation on the drift curve of the individual sensing device and the drift curve of the reference sensing device using the weight value S1 and the weight value C1 to generate a comprehensive drift curve. The service life analysis module includes a weight adjustment unit and a service life determination unit; the weight adjustment unit is used to set the weight value adjustment period and calculate the individual error H. single With reference error H crowd The individual weight value S is obtained after correction. 1，new Compared with the reference weight value C 1，new The lifespan determination unit is used to determine the lifespan based on the individual weight value S. 1，new Compared with the reference weight value C 1，new Calculate the overall drift curve of the next sensor of the same specification and model, and combine it with the drift limit value to obtain the service life of the sensor.

Images