Temperature calibration method and system of temperature sensor

Abstract

The invention discloses a temperature calibration method and system for a temperature sensor, and belongs to the technical field of temperature testing, and the method specifically comprises the steps: reading temperature data in real time through an intelligent temperature sensor, and transmitting the temperature data to a processing system; the processing system verifies and decodes the received data and then constructs a self-adaptive temperature reference model in combination with historical temperature data and current environment parameters; temperature deviation is identified, deviation reasons are analyzed through a machine learning algorithm, the temperature change trend is predicted, and correction parameters are automatically generated; the correction parameters are automatically distributed to the equipment end through the intelligent contract, so that instant adjustment and effect feedback of equipment setting are realized; the processing system collects feedback data, evaluates an adjustment effect, optimizes a self-adaptive temperature reference model and corrects a parameter generation strategy according to the adjustment effect, and forms a closed-loop optimization mechanism; according to the invention, the calibration efficiency and precision are improved, the environmental adaptability is enhanced, and the method is especially suitable for high-precision measurement requirements under extreme temperature conditions.

Description

Technical Field

[0001] The invention belongs to the technical field of temperature testing, and in particular to a temperature calibration method and system for a temperature sensor. Background Art

[0002] With the development of science and technology, temperature sensors have become increasingly important due to their wide applications in many fields such as environmental monitoring, industrial production, and medical health. Temperature sensors can convert temperature changes into electrical signal outputs and are key components for temperature measurement. However, in practical applications, due to differences in manufacturing processes, changes in the use environment, and other factors, the measurement results of temperature sensors often have certain errors, which directly affects the accuracy and reliability of their measurements.

[0003] Currently commonly used calibration methods include: zero-point calibration, linear calibration, and multi-point calibration. Among them, zero-point calibration usually measures the sensor once at a standard temperature and then adjusts the output based on the deviation; linear calibration calculates the linear equation of the sensor based on the measured values ​​at two different temperature points, which serves as the basis for subsequent temperature measurements; multi-point calibration measures at multiple known temperature points and corrects the sensor output by fitting the curve.

[0004] Although the above methods have improved the measurement accuracy of temperature sensors to a certain extent, there are still some shortcomings. For example, traditional calibration methods often rely on manual operations, which are time-consuming and inefficient; at the same time, these methods can usually only achieve good results under specific conditions. When environmental conditions change, the measurement error of the sensor may increase again. In addition, for use under extreme temperature conditions, the existing calibration methods may not meet the requirements of high-precision measurement.

[0005] For example, the patent with publication number CN114354006A discloses a device temperature testing method and a temperature testing system, including: setting a calibration temperature point and a verification temperature point, in the pre-calibration, controlling the temperature of the standard temperature source to the calibration temperature point and the verification temperature point, obtaining the temperature indications of the standard temperature sensor and the measuring temperature sensor, and then, in the device test, when the temperature of the device under test is stable at the calibration temperature point, obtaining the temperature indications of the measuring temperature sensor and the device under test, and then, in the post-calibration, controlling the temperature of the standard temperature source to the verification temperature point, obtaining the temperature indications of the standard temperature sensor and the measuring temperature sensor, calibrating the device under test according to the device test data results, and correcting and verifying the device test data according to the pre-calibration and post-calibration data. This technical solution can not only make the calibration result have sufficient credibility and reliability, but also make the calibration operation time as short as possible.

[0006] For example, the patent with publication number CN115165154A discloses a temperature sensor calibration system and method, including: first obtaining a first temperature measured by a standard temperature sensor and a second temperature measured by a sensor to be calibrated, then determining a calibration coefficient according to the first temperature and the second temperature, and finally calibrating the sensor to be calibrated according to the calibration coefficient. This technical solution embodiment calculates the calibration coefficient by measuring the first temperature and the second temperature, and then calibrates the temperature sensor according to the calibration coefficient, which can improve the calibration accuracy.

[0007] The above existing technologies all have the following problems: they rely on standard temperature sources and temperature sensors for calibration, and during the calibration process, the temperature of the standard temperature source needs to be controlled multiple times, and temperature indications need to be obtained at different temperature points, which increases the complexity of the calibration process; they lack adaptability and flexibility, and are sensitive to the environment; and the calibration process is relatively simple. Summary of the invention

[0008] In view of the shortcomings of the prior art, the present invention proposes a temperature calibration method and system for a temperature sensor, which reads temperature data in real time through an intelligent temperature sensor and transmits it to a processing system; after the processing system verifies and decodes the received data, it combines historical temperature data with current environmental parameters to build an adaptive temperature reference model; identifies temperature deviations, and analyzes the causes of the deviations through a machine learning algorithm, predicts temperature change trends, and automatically generates correction parameters; the correction parameters are automatically distributed to the device side through a smart contract, so as to realize instant adjustment of device settings and effect feedback; the processing system collects feedback data, evaluates the adjustment effect, and optimizes the adaptive temperature reference model and the correction parameter generation strategy accordingly, forming a closed-loop optimization mechanism; the present invention improves calibration efficiency and accuracy, enhances environmental adaptability, and is particularly suitable for high-precision measurement requirements under extreme temperature conditions.

[0009] To achieve the above object, the present invention provides the following technical solutions:

[0010] A temperature calibration method for a temperature sensor, comprising:

[0011] Step S1: Read the current temperature data from the intelligent temperature sensor, and transmit the read temperature data back to the processing system through the Internet of Things;

[0012] Step S2: The processing system verifies the received temperature data and decodes the verified temperature data to obtain a decoded temperature value;

[0013] Step S3: Establish an adaptive temperature reference model based on historical temperature data and current environmental parameters, compare the decoded temperature value with the standard temperature value output by the adaptive temperature reference model, and identify whether there is a deviation. If there is a deviation, use a machine learning algorithm to analyze the cause of the deviation, predict the temperature change trend, and automatically generate correction parameters based on the prediction results;

[0014] Step S4: automatically distributing the correction parameters to the device receiving end through the smart contract, and the device receiving end uses the received correction parameters to adjust the device settings and feeds back the adjustment effects to the processing system in real time;

[0015] Step S5: The processing system collects feedback data, compares and analyzes it with the expected effect, evaluates the adjustment effect by calculating the deviation rate, and adjusts the adaptive temperature reference model and correction parameter generation strategy according to the evaluation results to form a closed-loop optimization mechanism.

[0016] Specifically, the specific steps of step S3 include:

[0017] S3.1: Collect historical temperature data and current environmental parameter data, and pre-process the historical temperature data and current environmental parameter data to form fused data;

[0018] S3.2: Load the pre-built adaptive temperature benchmark model, and use the fusion data to train the pre-built adaptive temperature benchmark model;

[0019] S3.3: Acquire real-time temperature data, input the real-time temperature data into the trained adaptive temperature benchmark model, and obtain the standard temperature value hot1;

[0020] S3.4: Obtain the decoded temperature value hot2 in step S2, and compare the decoded temperature value hot2 with hot1.

[0021] Specifically, the specific steps of step S3 also include:

[0022] S3.5: Based on the comparison results,

[0023] If |hot1-hot2|=0, it means there is no deviation;

[0024] If |hot1-hot2|>0, it means there is a deviation. At the same time, load the machine learning model trained with the deviation data with known deviation causes, input hot2 into the trained machine learning model, and output the deviation prediction result and deviation cause classification;

[0025] S3.6: Based on historical temperature data and current environmental parameters, use time series analysis methods to predict temperature change trends, and automatically generate correction parameters based on the prediction results and deviation reasons;

[0026] S3.7: Apply the modified parameters to the actual system, verify the modified effect, and perform iterative optimization based on the feedback.

[0027] Specifically, the specific steps of S3.6 include:

[0028] S3.61: Obtain the decoded temperature value hot2, and form the decoded temperature data X={x1,...,x N}, and load the pre-trained time series analysis model, where x N represents the Nth decoded temperature data, where N represents the number of decoded temperature data;

[0029] S3.62: X={x1,...,x N} Input the pre-trained time series analysis model to obtain the temperature prediction result Y={y1,...,y N}, where y N Represents the prediction result of the Nth decoded temperature data;

[0030] S3.63: Use causal reasoning analysis to analyze the temperature prediction results Y = {y1,...,y N The cause of the deviation between the actual temperature and the actual temperature is analyzed, and the AutoML technology is used to automatically search and generate the optimal correction parameters based on the deviation cause analysis results;

[0031] S3.64: Apply the generated optimal correction parameters to the time series analysis model to adjust and optimize the time series analysis model.

[0032] Specifically, the specific formula for generating the optimal correction parameter in S3.63 is:

[0033]

[0034] in, represents the optimal correction parameter, R represents the AutoML search space, θ represents the time series analysis model parameter, hot 1,i represents the i-th standard temperature value, y i (θ) represents the i-th temperature data predicted using parameter θ, α represents the regularization term coefficient, and L1(θ) represents the L1 regularization term.

[0035] Specifically, the specific steps of step S5 include:

[0036] S5.1: The processing system collects the adjusted feedback data Z={z1,...,z N}, where z N Indicates the Nth adjusted feedback data;

[0037] S5.2: Set the deviation threshold h1 according to Z = {z1, ..., z N} and Y={y1,...,y N The degree of deviation between The adaptive temperature benchmark model is adjusted, where ε j Indicates the degree of deviation of the jth feedback data;

[0038] S5.3: Optimize the generation strategy of the correction parameters based on the deviation rate evaluation results and the adjustment of the adaptive temperature reference model.

[0039] A temperature calibration system for a temperature sensor, comprising: a data processing module, a data verification module, a deviation prediction module, a parameter adjustment module, and an effect evaluation module;

[0040] The data processing module is used to read the current temperature data from the intelligent temperature sensor in real time and transmit it back to the processing system;

[0041] The data verification module is used to verify and decode the received temperature data to obtain the temperature value after the temperature data is decoded;

[0042] The deviation prediction module is used to build an adaptive temperature benchmark model by using historical temperature data and real-time environmental parameters, monitor temperature deviations, analyze the root causes of deviations in combination with machine learning, and generate correction parameters for adjustment;

[0043] The parameter adjustment module is used to automatically distribute the correction parameters to the device receiving end through smart contract technology. The device receiving end adjusts the device settings according to the correction parameters and feeds back the adjustment effect to the processing system in real time;

[0044] The effect evaluation module is used to collect equipment adjustment feedback data and compare it with the expected effect, and optimize the temperature reference model and correction parameter generation strategy after evaluation to achieve closed-loop automatic tuning.

[0045] Specifically, the deviation prediction module includes: a model building unit, a deviation identification unit, and a prediction and correction unit;

[0046] The model building unit is used to build an adaptive temperature reference model using historical data and current environmental parameters;

[0047] The deviation identification unit is used to monitor the difference between the temperature data and the temperature reference model in real time and identify the temperature deviation;

[0048] The prediction and correction unit uses a machine learning algorithm to analyze the cause of the deviation, predict the temperature change trend, and generate correction parameters.

[0049] Specifically, the parameter adjustment module includes: a smart contract unit, a parameter distribution unit, and a feedback unit;

[0050] The smart contract unit is used to automatically formulate a parameter adjustment strategy according to the prediction results of the deviation prediction module;

[0051] The parameter distribution unit is used to send the correction parameters to the designated device receiving end;

[0052] The feedback unit is used to enable the device receiving end to receive parameters and adjust device settings, and at the same time feed back the adjustment effect to the processing system in real time.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] 1. The present invention proposes a temperature calibration system for a temperature sensor, and optimizes and improves the architecture, operation steps and processes. The system has the advantages of simple process, low investment and operation costs, and low production work costs.

[0055] 2. The present invention proposes a temperature calibration method for a temperature sensor, which realizes real-time transmission and processing of data through the Internet of Things. Combined with an adaptive temperature reference model and a machine learning algorithm, it can accurately identify and automatically correct temperature deviations, thereby ensuring the accuracy and reliability of the sensor measurement results. At the same time, its closed-loop optimization mechanism further improves the long-term stability and adaptability of the calibration system.

[0056] 3. The present invention proposes a temperature calibration method for a temperature sensor, which utilizes intelligent technologies, including the Internet of Things, adaptive modeling, machine learning, and smart contracts, to achieve efficient and intelligent calibration of the temperature sensor. It can not only correct deviations in real time, but also has self-learning capabilities to predict future changes, and forms a continuous optimization cycle through automated feedback and adjustment, thereby ensuring high-precision and long-term stable calibration effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 A schematic diagram of a temperature calibration method for a temperature sensor of the present invention;

[0058] Figure 2 A principle flow chart of a temperature calibration method for a temperature sensor of the present invention;

[0059] Figure 3A flow chart for generating correction parameters for a temperature calibration method of a temperature sensor according to the present invention;

[0060] Figure 4 This is a temperature calibration system architecture diagram of a temperature sensor of the present invention. DETAILED DESCRIPTION

[0061] Example 1

[0062] See also Figure 1 and Figure 2 , an embodiment of the present invention provides: a temperature calibration method for a temperature sensor, comprising the following steps:

[0063] Step S1: Read the current temperature data from the intelligent temperature sensor, and transmit the read temperature data back to the processing system through the Internet of Things;

[0064] Furthermore, the specific steps of step S1 include:

[0065] S1.1: Select high-precision smart temperature sensors and analyze their performance under different environmental conditions, including the impact of temperature, humidity, and pressure changes on sensor readings;

[0066] S1.2: Integrate an environmental adaptability algorithm into the sensor so that the sensor can be automatically calibrated through a built-in microprocessor. During the operation of the sensor, the environmental adaptability algorithm calculates calibration parameters in real time according to the current environmental conditions and applies them to the readings of the sensor to ensure the accuracy of the readings. However, it should be noted that the environmental adaptability algorithm is a prior art in the art and is not an inventive solution of the present application, and is not described in detail herein.

[0067] S1.3: After completing the calibration, the intelligent temperature sensor starts to read the temperature data of the current environment and transmits the read temperature data back to the processing system through the Internet of Things technology.

[0068] Step S2: The processing system verifies the received temperature data and decodes the verified temperature data to obtain a decoded temperature value;

[0069] Furthermore, the specific steps of step S2 include:

[0070] S2.1: The processing system receives raw temperature data from the temperature sensor through a communication interface, wherein the communication interface may be a serial port or a network;

[0071] S2.2: Calculate the received original temperature data using an XOR check algorithm to generate a check value, and compare the calculated check value with the check value attached to the data packet, wherein the formula of the XOR check algorithm is the prior art content in the art, is not the inventive solution of the present application, and is not described in detail here;

[0072] If the check values ​​are consistent, the data is considered complete and correct;

[0073] If the checksum values ​​are inconsistent, the system configuration determines whether to retry receiving, record an error, or issue an alarm;

[0074] S2.3: parsing the verified temperature data packet according to the communication protocol or data format specification, and extracting the temperature data portion;

[0075] S2.4: Convert the temperature data from the original binary format to a readable decimal or floating point format. The binary conversion of the data is the prior art content in this field, which is not the inventive solution of the present application and will not be elaborated here.

[0076] Step S3: Establish an adaptive temperature reference model based on historical temperature data and current environmental parameters, compare the decoded temperature value with the standard temperature value output by the adaptive temperature reference model, and identify whether there is a deviation. If there is a deviation, use a machine learning algorithm to analyze the cause of the deviation, predict the temperature change trend, and automatically generate correction parameters based on the prediction results;

[0077] Among them, environmental parameters include temperature, humidity and pressure.

[0078] Step S4: automatically distributing the correction parameters to the device receiving end through the smart contract, and the device receiving end uses the received correction parameters to adjust the device settings and feeds back the adjustment effects to the processing system in real time;

[0079] Furthermore, the specific steps of step S4 include:

[0080] S4.1: Design the logic of the smart contract, including the storage structure, distribution logic and trigger conditions of the correction parameters, and deploy the smart contract on the blockchain platform to ensure the security and immutability of the contract;

[0081] S4.2: Obtain the optimal correction parameters generated by causal reasoning and AutoML technology, and store the generated correction parameters in a smart contract, which can be a specific address or variable on the blockchain;

[0082] S4.3: Ensure that the receiving device has the ability to interact with the blockchain, including accessing the blockchain network, reading smart contract data, and implementing the corresponding interface or function on the receiving device to receive and process the correction parameters from the smart contract;

[0083] S4.4: Set trigger conditions, such as time intervals and external events. Once the conditions are met, the smart contract will automatically execute the distribution logic. At the same time, the smart contract will send the correction parameters to the designated device receiver according to the distribution logic;

[0084] S4.5: The device receiving end receives the correction parameters from the smart contract, and the device adjusts its own settings, such as temperature, humidity, and speed, according to the received correction parameters;

[0085] S4.6: The device receiving end feeds back the adjusted effect to the processing system in real time. The processing system receives and processes the information fed back by the device for subsequent analysis and optimization.

[0086] Step S5: The processing system collects feedback data, compares and analyzes it with the expected effect, evaluates the adjustment effect by calculating the deviation rate, and adjusts the adaptive temperature reference model and correction parameter generation strategy according to the evaluation results to form a closed-loop optimization mechanism.

[0087] Example 2

[0088] See also Figure 3 In this embodiment, the specific steps of step S3 include:

[0089] S3.1: Collect historical temperature data and current environmental parameter data, and pre-process the historical temperature data and current environmental parameter data to form fused data;

[0090] S3.2: Load the pre-built adaptive temperature benchmark model, and use the fusion data to train the pre-built adaptive temperature benchmark model;

[0091] S3.3: Acquire real-time temperature data, input the real-time temperature data into the trained adaptive temperature benchmark model, and obtain the standard temperature value hot1;

[0092] S3.4: Obtain the decoded temperature value hot2 in step S2, and compare the decoded temperature value hot2 with hot1;

[0093] S3.5: Based on the comparison results,

[0094] If |hot1-hot2|=0, it means there is no deviation;

[0095] If |hot1-hot2|>0, it means there is a deviation. At the same time, load the machine learning model trained with the deviation data with known deviation causes, input hot2 into the trained machine learning model, and output the deviation prediction result and deviation cause classification;

[0096] S3.6: Based on historical temperature data and current environmental parameters, use time series analysis methods to predict temperature change trends, and automatically generate correction parameters based on the prediction results and deviation reasons;

[0097] S3.7: Apply the modified parameters to the actual system, verify the modified effect, and perform iterative optimization based on the feedback.

[0098] The specific steps of S3.6 include:

[0099] S3.61: Obtain the decoded temperature value hot2, and form the decoded temperature data X={x1,...,x N}, and load the pre-trained time series analysis model, where x N represents the Nth decoded temperature data, where N represents the number of decoded temperature data;

[0100] S3.62: X={x1,...,x N} Input the pre-trained time series analysis model to obtain the temperature prediction result Y={y1,...,y N}, where y N represents the prediction result of the Nth decoded temperature data, wherein the construction and training process of the time series analysis model is the prior art content in this field, which is not the inventive solution of the present application and will not be described in detail here;

[0101] S3.63: Use causal reasoning analysis to analyze the temperature prediction results Y = {y1,...,y N The cause of the deviation between the actual temperature and the actual temperature is analyzed, and the AutoML technology is used to automatically search and generate the optimal correction parameters based on the deviation cause analysis results;

[0102] For example, if the temperature prediction result is too high, you can consider increasing the negative correction factor in the time series analysis model; if the temperature prediction result is too low, then increase the positive correction factor in the time series analysis model.

[0103] Furthermore, the specific steps of S3.63 include:

[0104] (1) Collect comparative data between temperature prediction results and actual temperature conditions, and preprocess the collected comparative data, including cleaning outliers, filling missing values, and normalizing data to ensure data quality;

[0105] (2) Identify the causal relationship that needs to be explored, that is, the potential causes between the temperature prediction deviation and the actual temperature situation;

[0106] (3) Use a causal graph to represent possible causal relationships, where nodes represent variables, such as predicted temperature, actual temperature, and environmental factors, and edges represent causal relationships between variables;

[0107] (4) Based on the causal graph, the potential causes of prediction deviation are analyzed using a causal forest-based machine learning algorithm to determine which factors are the main causes of prediction deviation;

[0108] (5) Quantitatively evaluate each potential cause to determine its contribution to the forecast deviation;

[0109] (6) Based on the results of the deviation cause analysis, the AutoML technology is used to automatically search for the optimal correction parameters. In the present invention, the AutoML technology uses an automated machine learning framework, and the automated machine learning framework is the prior art content in this field, and is not an inventive solution of the present application, and is not elaborated here.

[0110] Furthermore, the automated machine learning framework will automatically try different model structures, parameter configurations, etc. to find a model that minimizes prediction bias;

[0111] S3.64: Apply the generated optimal correction parameters to the time series analysis model to adjust and optimize the time series analysis model.

[0112] The specific formula for generating the optimal correction parameters in S3.63 is:

[0113]

[0114] in, represents the optimal correction parameter, R represents the AutoML search space, θ represents the time series analysis model parameter, hot 1,i represents the i-th standard temperature value, y i (θ) represents the i-th temperature data predicted using parameter θ, α represents the regularization term coefficient, and L1(θ) represents the L1 regularization term.

[0115] It should be noted that the formula in the present invention helps the time series analysis model to better fit all data points by adding additional parameters or variables, thereby further improving the flexibility and prediction accuracy of the time series analysis model.

[0116] The specific steps of step S5 include:

[0117] S5.1: The processing system collects the adjusted feedback data Z={z1,...,z N}, where z N Indicates the Nth adjusted feedback data;

[0118] S5.2: Set the deviation threshold h1 according to Z = {z1, ..., z N} and Y={y1,...,y N The degree of deviation between The adaptive temperature benchmark model is adjusted, where ε j Indicates the degree of deviation of the jth feedback data;

[0119] S5.3: Optimize the generation strategy of the correction parameters based on the deviation rate evaluation results and the adjustment of the adaptive temperature reference model.

[0120] Example 3

[0121] See also Figure 4 Another embodiment provided by the present invention is a temperature calibration system for a temperature sensor, comprising:

[0122] Data processing module, data verification module, deviation prediction module, parameter adjustment module, effect evaluation module;

[0123] The data processing module is used to read the current temperature data from the intelligent temperature sensor in real time and transmit the temperature data back to the processing system safely and quickly through the Internet of Things technology;

[0124] The data verification module is used to verify and decode the received temperature data to obtain an accurate temperature value;

[0125] The deviation prediction module is used to build an adaptive temperature benchmark model using historical temperature data and real-time environmental parameters. It monitors temperature deviations, analyzes the root causes of deviations through machine learning, and generates correction parameters for adjustment in a timely manner.

[0126] The parameter adjustment module is used to automatically distribute the correction parameters to the device receiving end through smart contract technology. The device receiving end adjusts the device settings according to the correction parameters and feeds back the adjustment effects to the processing system in real time;

[0127] The effect evaluation module is used to collect equipment adjustment feedback data and compare it with the expected effect. After the evaluation, it optimizes the temperature benchmark model and the correction parameter generation strategy to achieve closed-loop automatic tuning.

[0128] The deviation prediction module includes: a model building unit, a deviation identification unit, and a prediction and correction unit;

[0129] A model building unit, used to build an adaptive temperature benchmark model using historical data and current environmental parameters;

[0130] A deviation identification unit is used to monitor the difference between the temperature data and the temperature reference model in real time and identify the temperature deviation;

[0131] The prediction and correction unit uses machine learning algorithms to analyze the causes of deviations, predict temperature change trends, and generate correction parameters.

[0132] The parameter adjustment module includes: smart contract unit, parameter distribution unit, and feedback unit;

[0133] Smart contract unit, used to automatically formulate parameter adjustment strategies based on the prediction results of the deviation prediction module;

[0134] A parameter distribution unit is used to send the modified parameters to the designated device receiving end to ensure that the device can work according to the new parameters;

[0135] The feedback unit is used to enable the device receiving end to receive parameters and adjust device settings, and at the same time feed back the adjustment effects to the processing system in real time.

[0136] The effect evaluation module includes: data analysis unit, effect evaluation unit, and optimization unit;

[0137] The data analysis unit is used to collect data fed back by the equipment, conduct in-depth analysis on the collected adjustment effect data, and evaluate the actual effect of parameter adjustment, including comparing the temperature data before and after adjustment and analyzing whether the adjustment effect meets expectations;

[0138] The effect evaluation unit is used to evaluate the effect of the adjustment by calculating the deviation rate index;

[0139] The optimization unit is used to adjust and optimize the adaptive temperature benchmark model and the correction parameter generation strategy according to the evaluation results.

[0140] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in the field may also change, modify, replace and modify the above-mentioned embodiments without departing from the purpose and scope of protection of the present invention, and all of these are within the protection of the present invention.

Claims

1. A temperature calibration method for a temperature sensor, characterized in that: include: Step S1: Read the current temperature data from the intelligent temperature sensor, and transmit the read temperature data back to the processing system through the Internet of Things; Step S2: The processing system verifies the received temperature data and decodes the verified temperature data to obtain a decoded temperature value; Step S3: Establish an adaptive temperature reference model based on historical temperature data and current environmental parameters, compare the decoded temperature value with the standard temperature value output by the adaptive temperature reference model, and identify whether there is a deviation. If there is a deviation, use a machine learning algorithm to analyze the cause of the deviation, predict the temperature change trend, and automatically generate correction parameters based on the prediction results; Step S4: automatically distributing the correction parameters to the device receiving end through the smart contract, and the device receiving end uses the received correction parameters to adjust the device settings and feeds back the adjustment effects to the processing system in real time; Step S5: The processing system collects feedback data, compares and analyzes it with the expected effect, evaluates the adjustment effect by calculating the deviation rate, and adjusts the adaptive temperature reference model and correction parameter generation strategy according to the evaluation results to form a closed-loop optimization mechanism.

2. A temperature calibration method for a temperature sensor as claimed in claim 1, characterized in that: The specific steps of step S3 include: S3.1: Collect historical temperature data and current environmental parameter data, and pre-process the historical temperature data and current environmental parameter data to form fused data; S3.2: Load the pre-built adaptive temperature benchmark model, and use the fusion data to train the pre-built adaptive temperature benchmark model; S3.3: Acquire real-time temperature data, input the real-time temperature data into the trained adaptive temperature benchmark model, and obtain the standard temperature value hot1; S3.4: Obtain the decoded temperature value hot2 in step S2, and compare the decoded temperature value hot2 with hot1.

3. A temperature calibration method for a temperature sensor as claimed in claim 2, characterized in that: The specific steps of step S3 also include: S3.5: Based on the comparison results, If |hot1-hot2|=0, it means there is no deviation; If |hot1-hot2|>0, it means there is a deviation. At the same time, load the machine learning model trained with the deviation data with known deviation causes, input hot2 into the trained machine learning model, and output the deviation prediction result and deviation cause classification; S3.6: Based on historical temperature data and current environmental parameters, use time series analysis methods to predict temperature change trends, and automatically generate correction parameters based on the prediction results and deviation reasons; S3.7: Apply the modified parameters to the actual system, verify the modified effect, and perform iterative optimization based on the feedback.

4. A temperature calibration method for a temperature sensor as claimed in claim 3, characterized in that: The specific steps of S3.6 include: S3.61: Obtain the decoded temperature value hot2, and form the decoded temperature data X={x1,…,x N }, and load the pre-trained time series analysis model, where x N represents the Nth decoded temperature data, where N represents the number of decoded temperature data; S3.62: X={x1,…,x N } Input the pre-trained time series analysis model to obtain the temperature prediction result Y={y1,…,y N }, where y N Represents the prediction result of the Nth decoded temperature data; S3.63: Use causal reasoning analysis to analyze the temperature prediction results Y = {y1,...,y N The cause of the deviation between the actual temperature and the actual temperature is analyzed, and the AutoML technology is used to automatically search and generate the optimal correction parameters based on the deviation cause analysis results; S3.64: Apply the generated optimal correction parameters to the time series analysis model to adjust and optimize the time series analysis model.

5. A temperature calibration method for a temperature sensor as claimed in claim 4, characterized in that: The specific formula for generating the optimal correction parameter in S3.63 is: in, represents the optimal correction parameter, R represents the AutoML search space, θ represents the time series analysis model parameter, hot 1,i represents the i-th standard temperature value, y i (θ) represents the i-th temperature data predicted using parameter θ, α represents the regularization term coefficient, and L1(θ) represents the L1 regularization term.

6. A temperature calibration method for a temperature sensor as claimed in claim 5, characterized in that: The specific steps of step S5 include: S5.1: The processing system collects the adjusted feedback data Z={z1,...,z N }, where z N Indicates the Nth adjusted feedback data; S5.2: Set the deviation threshold h1 according to Z = {z1, ..., z N } and Y={y1,...,y N The degree of deviation between The adaptive temperature benchmark model is adjusted, where ε j Indicates the degree of deviation of the jth feedback data; S5.3: Optimize the generation strategy of the correction parameters based on the deviation rate evaluation results and the adjustment of the adaptive temperature reference model.

7. A temperature calibration system for a temperature sensor, used to implement a temperature calibration method for a temperature sensor according to any one of claims 1 to 6, characterized in that: include: Data processing module, data verification module, deviation prediction module, parameter adjustment module, effect evaluation module; The data processing module is used to read the current temperature data from the intelligent temperature sensor in real time and transmit it back to the processing system; The data verification module is used to verify and decode the received temperature data to obtain the temperature value after the temperature data is decoded; The deviation prediction module is used to build an adaptive temperature benchmark model by using historical temperature data and real-time environmental parameters, monitor temperature deviations, analyze the root causes of deviations in combination with machine learning, and generate correction parameters for adjustment; The parameter adjustment module is used to automatically distribute the correction parameters to the device receiving end through smart contract technology. The device receiving end adjusts the device settings according to the correction parameters and feeds back the adjustment effect to the processing system in real time; The effect evaluation module is used to collect equipment adjustment feedback data and compare it with the expected effect, and optimize the temperature reference model and correction parameter generation strategy after evaluation to achieve closed-loop automatic tuning.

8. A temperature calibration system for a temperature sensor as claimed in claim 7, characterized in that: The deviation prediction module includes: a model building unit, a deviation identification unit, and a prediction and correction unit; The model building unit is used to build an adaptive temperature reference model using historical data and current environmental parameters; The deviation identification unit is used to monitor the difference between the temperature data and the temperature reference model in real time and identify the temperature deviation; The prediction and correction unit uses a machine learning algorithm to analyze the cause of the deviation, predict the temperature change trend, and generate correction parameters.

9. A temperature calibration system for a temperature sensor as claimed in claim 8, characterized in that: The parameter adjustment module includes: a smart contract unit, a parameter distribution unit, and a feedback unit; The smart contract unit is used to automatically formulate a parameter adjustment strategy according to the prediction results of the deviation prediction module; The parameter distribution unit is used to send the correction parameters to the designated device receiving end; The feedback unit is used to enable the device receiving end to receive parameters and adjust device settings, and at the same time feed back the adjustment effect to the processing system in real time.

Images