Intelligent adaptive calibration method and calibration system for sensors

Abstract

The present invention relates to the technical field of sensor calibration, and discloses an intelligent adaptive calibration method and calibration system for sensors. The adaptive calibration method is applied to a sensor monitoring unit, and specifically includes the following steps: S101: receiving a calibration request of a sensor, wherein the calibration request includes the time series, sensor spectrum, and sensor wavelet transform state of the sensor within a stage time. The present invention collects the time series, sensor spectrum, and sensor wavelet transform state data of the sensor, first realizes the transformation based on the correction method of the data, and then uses a preset mathematical model to predict the matching options of the sensor, and detects whether there is a matching request corresponding to the actual output value of the specified matching option, determines the actual output value of the sensor in the matching request and compares it with the predicted output value of the sensor, thereby realizing accurate duplicate checking and calibration, improving the accuracy of the calibration value, and ensuring the accuracy of the subsequent data collected by the sensor.

Description

Technical Field

[0001] The present invention relates to the technical field of sensor calibration, and in particular to an intelligent adaptive calibration method and a calibration system for a sensor. Background Art

[0002] Sensors are widely used in industrial automation technology, detection technology, information processing and other fields. Sensors can be directly used to measure current, magnetic induction intensity, magnetic field direction (angle), etc.

[0003] However, since the sensor itself has hardware errors and other hardware systems used with it also have hardware errors, parameter calibration is generally required in practical applications to ensure detection accuracy. Traditional calibration methods have disadvantages such as expensive calibration tooling, risk of matching errors, long time required for overall parameter calibration, many manual steps, and low overall efficiency. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent adaptive calibration method and calibration system for sensors, determine the actual output value of the sensor in the matching request and compare it with the predicted output value of the sensor, thereby achieving accurate duplicate checking and calibration, improving the accuracy of the calibration value, and aiming to solve the problems in the prior art.

[0005] The present invention is implemented as follows: an intelligent adaptive calibration method for a sensor is applied to a sensor monitoring unit and specifically comprises the following steps:

[0006] S101: Receive a calibration request for a sensor, wherein the calibration request includes a time series, a spectrum, and a wavelet transform state of the sensor within a stage time period;

[0007] S102: Acquire spectrum state and wavelet transform state data of the sensor in a time series, input the spectrum state and wavelet transform state data of the sensor in the time series into a preset data classification model, and obtain a classification result output based on the spectrum state and wavelet transform state data of the sensor in the time series;

[0008] S103: Obtain classification results of the sensor's spectrum state and wavelet transform state data output in the time series, place the classification results in an adjustment unit to complete data comparison, and obtain a reference feature data deviation error;

[0009] S104: placing the reference feature data offset error in an adjustment unit to perform sensor offset adjustment, and obtaining an offset identifier. After the sensor offset adjustment, obtaining the adjusted spectral state and wavelet transform state data of the sensor in the time series again;

[0010] S105: Identifying the adjusted spectrum state and wavelet transform state data of the sensor in the time series, and placing them in a preset mathematical model, predicting matching options of the sensor using the preset mathematical model, and detecting whether there is a matching request corresponding to the actual output value of the specified matching option;

[0011] S106: If the specified matching option has a corresponding matching request, the actual output value of the sensor in the matching request is compared with the predicted output value of the sensor. If the calibration condition is met, the sensor data is compensated and corrected again according to the current actual output value.

[0012] Furthermore, in S101, receiving a calibration request of a sensor includes:

[0013] receiving a connection request sent by a sensor through a preset network, wherein the connection request is used to request to establish a connection with the sensor monitoring unit;

[0014] detecting a sending path of the connection request and determining characteristic parameters of the sensor required for calibration;

[0015] After obtaining the transmission path of the sensor, the sensor is connected according to the connection request. After the connection is completed, a calibration request of the sensor is received.

[0016] Furthermore, the calibration request includes a time series, a sensor spectrum, and a wavelet transform state of the sensor within a stage time, wherein the stage time includes:

[0017] The time series data sampling of the sensor is retained for at least two periodic segments;

[0018] The sensor spectrum is intercepted into at least multiple frequency bands, and the multiple frequency bands are a set of point differences recorded on the spectrum of the sensor's time series data;

[0019] The sensor wavelet transform state at least intercepts the requirements of gradually performing multi-scale refinement on the signal to achieve time subdivision at high frequencies, frequency subdivision at low frequencies, and automatic adaptation to time-frequency signal analysis, and can focus on any details of the signal.

[0020] Furthermore, the preset network includes one or a combination of 3G network, 4G network, 5G network, and WIFI network.

[0021] Furthermore, in S102, the spectrum state and wavelet transform state data of the sensor in the time series are obtained, and the spectrum state and wavelet transform state data of the sensor in the time series are input into a preset data classification model, including:

[0022] Acquire multiple data classification models, and sort the types of the multiple data classification models in order;

[0023] The content text of the spectrum state and wavelet transform state data of the sensor in the time series is obtained, and the spectrum state and wavelet transform state data of the sensor in the time series are input into the data classification model that is ranked first.

[0024] Furthermore, in S104, the reference characteristic data deviation error is placed in the adjustment unit to perform sensor deviation adjustment, and a deviation identifier is obtained, including:

[0025] The adjustment unit obtains a reference feature data deviation error based on the classification result of the data output, and then identifies the feature data parameters of the current sensor;

[0026] The reference feature data deviation error is superimposed on the identified feature data parameter, and the reference feature data deviation error is provided with a direction feature. The adjustment unit first identifies the direction feature and then determines the numerical value of the feature data parameter to adjust the direction.

[0027] Furthermore, in S105, a preset mathematical model is used to predict the matching options of the sensor, and a matching request corresponding to the actual output value of the specified matching option is detected, including:

[0028] Establish a mathematical model to describe the error characteristics of the sensor, and then use the mathematical model to predict the parameter output value of the sensor;

[0029] Obtain the parameter output values ​​of the sensor predicted by the mathematical model, and arrange the parameter output values ​​according to the preset parameter arrangement method to form matching options;

[0030] Detecting the spectrum state of the sensor in the time series and the actual output value path of the wavelet transform state data and the integrity of the data after adjustment;

[0031] If it is checked that the actual output value path and data of the sensor's spectrum state and wavelet transform state data in the time series are complete, then a matching option is specified to establish a corresponding matching request.

[0032] Furthermore, in S106, if the specified matching option does not correspond to the matching request, a feedback signal is output to the adjustment unit to determine whether the actual output value path and data of the spectral state and wavelet transform state data of the sensor in the time series after adjustment are incomplete;

[0033] If the actual output numerical path and data of the spectrum state and wavelet transform state data of the adjusted sensor in the time series are incomplete, the corresponding matching request will continue to be requested within the cycle time until the matching request corresponding to the specified matching option is completed.

[0034] Compared with the prior art, the intelligent adaptive sensor calibration method and calibration system provided by the present invention have the following beneficial effects:

[0035] 1. By collecting the sensor's time series, sensor spectrum, and sensor wavelet transform state data, the transformation is first implemented based on the data correction method, the reference feature data offset error is placed in the adjustment unit to perform sensor offset adjustment, and the offset identifier is obtained, and the content of the adjustment data is recorded. Then, through the system's preset functional parameters, a preset mathematical model is generated through the preset functional parameters. The preset mathematical model is used to predict the sensor's matching options, and it is detected whether the specified matching options have a matching request corresponding to the actual output value. The actual output value of the sensor in the matching request is determined and compared with the predicted sensor output value, thereby achieving accurate duplicate detection and calibration, improving the accuracy of the calibration value, and ensuring the accuracy of the sensor's subsequent data collection;

[0036] 2. The time series data sampling of the sensor retains at least two periodic segments; the sensor spectrum is intercepted with at least multiple frequency bands, and the multiple frequency bands are the set of point differences recorded by the time series data of the sensor on the spectrum; the sensor wavelet transform state at least intercepts the signal to gradually perform multi-scale refinement to achieve the requirements of time subdivision at high frequencies, frequency subdivision at low frequencies, and automatic adaptation to time-frequency signal analysis, and can focus on any details of the signal, so that when performing the data comparison, the data collection deviation can be repeated multiple times to maximize the acquisition of the benchmark feature data deviation error, and then the benchmark feature data deviation error is placed in the adjustment unit for sensor deviation adjustment, so as to obtain the optimal data deviation error and improve the sensor monitoring accuracy.

[0037] An intelligent adaptive calibration system for a sensor, applied to the above-mentioned intelligent adaptive calibration method, comprises:

[0038] A receiving and identifying module is configured to receive a calibration request from a sensor, input the spectral state and wavelet transform state data of the sensor in a time series into a preset data classification model, and obtain a classification result based on the spectral state and wavelet transform state data output by the sensor in the time series;

[0039] The data analysis module is used to classify the sensor's spectrum state and wavelet transform state data output in the time series, place the classification results in the adjustment unit to complete data comparison, and obtain the deviation error of the benchmark feature data;

[0040] A correction acquisition module is used to place the deviation error of the reference characteristic data into the adjustment unit to adjust the sensor deviation and obtain the deviation mark;

[0041] a model matching module, configured to identify the adjusted spectral state and wavelet transform state data of the sensor in a time series, place the data in a preset mathematical model, and predict matching options of the sensor using the preset mathematical model;

[0042] A prediction comparison module, configured to determine whether a matching option corresponds to the matching request, and if so, to compare the actual output value of the sensor in the matching request with the predicted output value of the sensor;

[0043] The compensation module is used to compensate and correct the sensor data again according to the current actual output value.

[0044] Specifically, the model matching module includes:

[0045] A model generation unit is used to obtain the system's preset functional parameters and generate a preset mathematical model based on the preset functional parameters;

[0046] An output matching unit, configured to predict the matching options of the sensor using a preset mathematical model and detect whether a matching request corresponding to an actual output value exists for the specified matching option;

[0047] The request acquisition unit is configured to determine whether a matching request corresponds to a specified matching option, and compare the actual output value of the sensor in the matching request with the predicted output value of the sensor. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a schematic flow chart of the intelligent adaptive calibration method for sensors proposed in the present invention;

[0049] Figure 2 This is a schematic block diagram of the process of receiving a sensor calibration request in the intelligent adaptive sensor calibration method proposed by the present invention;

[0050] Figure 3 This is a schematic diagram of the structure of the intelligent adaptive calibration system for sensors proposed in the present invention;

[0051] Figure 4 This is a structural diagram of the model matching module in the intelligent adaptive calibration system for sensors proposed in the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0053] The implementation of the present invention is described in detail below with reference to specific embodiments.

[0054] The same or similar numbers in the drawings of this embodiment correspond to the same or similar parts; in the description of the present invention, it should be understood that if the terms "up", "down", "left", "right", etc. indicate directions or positional relationships, they are based on the directions or positional relationships shown in the drawings. This is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes and cannot be understood as limiting the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0055] Reference Figure 1-2 As shown, the intelligent adaptive calibration method of the sensor is applied to the sensor monitoring unit and specifically includes the following steps:

[0056] S101: Receive a calibration request for a sensor, where the calibration request includes a time series, a sensor spectrum, and a wavelet transform state of the sensor within a stage time period;

[0057] Receiving a calibration request from a sensor includes:

[0058] receiving a connection request sent by a sensor through a preset network, the connection request being used to request establishment of a connection with a sensor monitoring unit;

[0059] Detecting the sending path of the connection request and determining the characteristic parameters of the sensor required for calibration;

[0060] After obtaining the sensor's transmission path, it connects to the sensor according to the connection request. After the connection is completed, it receives the sensor's calibration request.

[0061] S102: Acquire spectrum state and wavelet transform state data of the sensor in the time series, input the spectrum state and wavelet transform state data of the sensor in the time series into a preset data classification model, and obtain a classification result based on the spectrum state and wavelet transform state data output of the sensor in the time series;

[0062] The process of obtaining the spectrum state and wavelet transform state data of the sensor in the time series and inputting the spectrum state and wavelet transform state data of the sensor in the time series into a preset data classification model includes:

[0063] Obtain multiple data classification models, and sort the types of the multiple data classification models in order;

[0064] Obtaining content text of the sensor's spectrum state and wavelet transform state data in a time series, and inputting the sensor's spectrum state and wavelet transform state data in the time series into a data classification model ranked first;

[0065] S103: Obtain classification results of the sensor's spectrum state and wavelet transform state data output in the time series, place the classification results in an adjustment unit to complete data comparison, and obtain the reference feature data deviation error;

[0066] S104: placing the reference feature data offset error in an adjustment unit to perform sensor offset adjustment, and obtaining an offset identifier. After the sensor offset adjustment, obtaining the adjusted sensor's spectrum state and wavelet transform state data in a time series again;

[0067] The process of placing the deviation error of the reference characteristic data in the adjustment unit to adjust the sensor deviation and obtaining the deviation identifier includes:

[0068] The adjustment unit obtains the reference feature data deviation error based on the classification result of the data output, and then identifies the feature data parameters of the current sensor;

[0069] The reference characteristic data deviation error is superimposed on the identified characteristic data parameter, and a direction feature is set on the reference characteristic data deviation error. The adjustment unit first identifies the direction feature and then determines the numerical value of the characteristic data parameter to adjust the direction;

[0070] S105: Identifying the spectral state and wavelet transform state data of the adjusted sensor in the time series, and placing them in a preset mathematical model, using the preset mathematical model to predict the matching options of the sensor, and detecting whether there is a matching request corresponding to the actual output value of the specified matching option;

[0071] The method of using a preset mathematical model to predict the matching options of the sensor and detecting whether there is a matching request corresponding to the actual output value of the specified matching option includes:

[0072] Establish a mathematical model to describe the error characteristics of the sensor, and then use the mathematical model to predict the parameter output value of the sensor;

[0073] Obtain the parameter output values ​​of the sensor predicted by the mathematical model, and arrange the parameter output values ​​according to the preset parameter arrangement method to form matching options;

[0074] Detect the actual output value path and data integrity of the spectral state of the adjusted sensor in the time series and the wavelet transform state data;

[0075] If the actual output numerical path and data of the sensor's spectrum state and wavelet transform state data in the time series are checked to be complete, then a matching option is specified to establish a corresponding matching request;

[0076] S106: If there is a corresponding matching request for the specified matching option, the actual output value of the sensor in the matching request is determined to be compared with the predicted output value of the sensor. If the calibration condition is met, the sensor data is compensated and corrected again according to the current actual output value. The time series, sensor spectrum and wavelet transform state data of the sensor are collected. First, the transformation is tried based on the data correction method. The deviation error of the benchmark feature data is placed in the adjustment unit to adjust the sensor deviation, and the deviation identifier is obtained. The content of the adjustment data is recorded. Then, the system preset functional parameters are used to generate a preset mathematical model through the preset functional parameters. The preset mathematical model is used to predict the matching options of the sensor, and it is detected whether there is a matching request corresponding to the actual output value of the specified matching option. The actual output value of the sensor in the matching request is determined to be compared with the predicted output value of the sensor, thereby achieving accurate duplicate detection and calibration, improving the accuracy of the calibration value, and ensuring the accuracy of the subsequent data collected by the sensor.

[0077] Specifically, model-based sensor autocalibration uses a model to automatically calibrate the sensor. The core of this approach is to compare the sensor's output with the actual value, thereby adjusting the model parameters to achieve autocalibration. Commonly used models include linear models, nonlinear models, and neural network models.

[0078] In this embodiment, the calibration request includes the time series, sensor spectrum, and sensor wavelet transform state of the sensor within a phase time. The phase time includes:

[0079] The time series data sampling of the sensor is retained for at least two period segments;

[0080] The sensor spectrum is intercepted with at least multiple frequency bands, which are a set of point differences recorded on the spectrum of the sensor's time series data. This allows for repeated data collection deviations during data comparison to maximize the acquisition of the baseline feature data deviation error. The baseline feature data deviation error is then placed in an adjustment unit to perform sensor deviation adjustment, thereby obtaining the optimal data deviation error and improving the sensor monitoring accuracy.

[0081] The sensor wavelet transform state at least intercepts the signal to gradually perform multi-scale refinement to meet the requirements of time subdivision at high frequencies, frequency subdivision at low frequencies, and automatic adaptation to time-frequency signal analysis, and can focus on any details of the signal.

[0082] In this embodiment, the preset network includes one of a 3G network, a 4G network, a 5G network, and a WIFI network, or a combination thereof.

[0083] In S106 of this embodiment, if the specified matching option does not have a corresponding matching request, a feedback signal is output to the adjustment unit to determine whether the actual output value path and data of the spectral state and wavelet transform state data of the adjusted sensor in the time series are incomplete;

[0084] If the actual output numerical path of the spectrum state and wavelet transform state data of the adjusted sensor in the time series is incomplete, the corresponding matching request will continue to be requested within the cycle time until the matching request corresponding to the specified matching option is completed.

[0085] Reference Figure 3-4 As shown, the intelligent adaptive calibration system for sensors is applied to the above-mentioned intelligent adaptive calibration method. The system includes: a receiving and identifying module for receiving a calibration request from a sensor, and inputting the spectral state and wavelet transform state data of the sensor in a time series into a preset data classification model to obtain a classification result based on the spectral state and wavelet transform state data output by the sensor in the time series; a data analysis module for classifying the spectral state and wavelet transform state data output by the sensor in the time series, placing the classification result in an adjustment unit for data comparison, and obtaining a reference feature data offset error; a correction acquisition module for placing the reference feature data offset error in the adjustment unit for sensor offset adjustment and obtaining a offset identifier; a model matching module for identifying the adjusted spectral state and wavelet transform state data of the sensor in the time series, placing the adjusted data in a preset mathematical model, and using the preset mathematical model to predict matching options for the sensor; a prediction and comparison module for determining whether a matching request exists for the matching option, and if so, comparing the actual sensor output value in the matching request with the predicted sensor output value; and a compensation module for further compensating and correcting the sensor data based on the current actual output value.

[0086] Specifically, the time series data sampling of the sensor retains at least two periodic segments; the sensor spectrum is intercepted with at least multiple frequency bands, and the multiple frequency bands are a set of point differences recorded by the time series data of the sensor on the spectrum; the wavelet transform state of the sensor is at least intercepted with the requirements of gradually performing multi-scale refinement of the signal to achieve time subdivision at high frequencies, frequency subdivision at low frequencies, and automatic adaptation to time-frequency signal analysis, and can focus on any details of the signal, so that when performing the data comparison, the data collection deviation can be repeated multiple times to maximize the acquisition of the benchmark feature data deviation error, and then the benchmark feature data deviation error is placed in the adjustment unit for sensor deviation adjustment, so as to obtain the optimal data deviation error and improve the sensor monitoring accuracy.

[0087] In this embodiment, the model matching module includes: a model generation unit, configured to obtain system preset functional parameters and generate a preset mathematical model based on the preset functional parameters; an output matching unit, configured to predict sensor matching options using the preset mathematical model and detect whether a matching request corresponding to an actual output value exists for the specified matching option; and a request acquisition unit, configured to determine whether a matching request exists for the specified matching option and compare the actual sensor output value in the matching request with the predicted sensor output value. By collecting sensor time series, sensor spectrum, and sensor wavelet transform state data, a transformation is first performed based on a data correction method. The offset error of the reference feature data is placed in an adjustment unit for sensor offset adjustment, an offset identifier is obtained, and the content of the adjustment data is recorded. Then, the preset mathematical model is generated based on the preset functional parameters, and the matching options of the sensor are predicted using the preset mathematical model. The matching options of the sensor are then detected. A matching request corresponding to the actual output value of the specified matching option is detected. The actual sensor output value in the matching request is compared with the predicted sensor output value, thereby achieving accurate duplicate detection and calibration, improving the accuracy of the calibration value, and ensuring the accuracy of subsequent data collected by the sensor.

[0088] For the sensor wavelet transform state, it corresponds to a signal with limited energy, and its amplitude is in the form of alternating positive and negative oscillations. Wavelet transform is a localized analysis of time and frequency. It gradually refines the signal at multiple scales through scaling and translation operations, ultimately achieving time subdivision at high frequencies and frequency subdivision at low frequencies. It can automatically adapt to the requirements of time-frequency signal analysis, thereby focusing on any details of the signal.

[0089] The process of wavelet transform:

[0090] ① Select the wavelet basis function and initialize the scale factor and translation factor;

[0091] ②Multiply and integrate the signal and the wavelet to obtain the wavelet coefficient at the corresponding time;

[0092] ③ Move the translation factor until the end of the signal;

[0093] ④ Change the scale factor and repeat steps 2 and 3 to complete the wavelet transformation.

[0094] In this embodiment, the entire operation process can be controlled by a computer, and in each operation link, sensors can be set to provide signal feedback to achieve the sequential execution of steps. These are all common knowledge of current automated control and will not be described in detail in this embodiment.

[0095] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An intelligent adaptive calibration method for a sensor, characterized in that: Applied to the sensor monitoring unit, specifically including the following steps: S101: Receive a calibration request for a sensor, wherein the calibration request includes a time series, a spectrum, and a wavelet transform state of the sensor within a stage time period; S102: Acquire spectrum state and wavelet transform state data of the sensor in a time series, input the spectrum state and wavelet transform state data of the sensor in the time series into a preset data classification model, and obtain a classification result output based on the spectrum state and wavelet transform state data of the sensor in the time series; S103: Obtain classification results of the sensor's spectrum state and wavelet transform state data output in the time series, place the classification results in an adjustment unit to complete data comparison, and obtain a reference feature data deviation error; S104: placing the reference feature data offset error in an adjustment unit to perform sensor offset adjustment, and obtaining an offset identifier. After the sensor offset adjustment, obtaining the adjusted spectral state and wavelet transform state data of the sensor in the time series again; S105: Identifying the adjusted spectrum state and wavelet transform state data of the sensor in the time series, and placing them in a preset mathematical model, predicting matching options of the sensor using the preset mathematical model, and detecting whether there is a matching request corresponding to the actual output value of the specified matching option; S106: If the specified matching option has a corresponding matching request, the actual output value of the sensor in the matching request is compared with the predicted output value of the sensor. If the calibration condition is met, the sensor data is compensated and corrected again according to the current actual output value.

2. The intelligent adaptive calibration method for a sensor according to claim 1, wherein: In S101, a calibration request of a sensor is received, including: receiving a connection request sent by a sensor through a preset network, wherein the connection request is used to request to establish a connection with the sensor monitoring unit; detecting a sending path of the connection request and determining characteristic parameters of the sensor required for calibration; After obtaining the transmission path of the sensor, the sensor is connected according to the connection request. After the connection is completed, a calibration request of the sensor is received.

3. The intelligent adaptive calibration method for a sensor according to claim 2, wherein: The calibration request includes the time series, sensor spectrum, and sensor wavelet transform state of the sensor within a stage time, wherein the stage time includes: The time series data sampling of the sensor is retained for at least two periodic segments; The sensor spectrum is intercepted into at least multiple frequency bands, and the multiple frequency bands are a set of point differences recorded on the spectrum of the sensor's time series data; The sensor wavelet transform state at least intercepts the requirements of gradually performing multi-scale refinement on the signal to achieve time subdivision at high frequencies, frequency subdivision at low frequencies, and automatic adaptation to time-frequency signal analysis, and can focus on any details of the signal.

4. The intelligent adaptive calibration method for a sensor according to claim 3, wherein: The preset network includes one of 3G network, 4G network, 5G network, WIFI network or a combination thereof.

5. The intelligent adaptive calibration method for a sensor according to claim 4, wherein: In S102, the spectrum state and wavelet transform state data of the sensor in the time series are obtained, and the spectrum state and wavelet transform state data of the sensor in the time series are input into a preset data classification model, including: Acquire multiple data classification models, and sort the types of the multiple data classification models in order; The content text of the spectrum state and wavelet transform state data of the sensor in the time series is obtained, and the spectrum state and wavelet transform state data of the sensor in the time series are input into the data classification model that is ranked first.

6. The intelligent adaptive calibration method for a sensor according to claim 5, wherein: In S104, the reference characteristic data deviation error is placed in the adjustment unit to perform sensor deviation adjustment, and a deviation identifier is obtained, including: The adjustment unit obtains a reference feature data deviation error based on the classification result of the data output, and then identifies the feature data parameters of the current sensor; The reference feature data deviation error is superimposed on the identified feature data parameter, and the reference feature data deviation error is provided with a direction feature. The adjustment unit first identifies the direction feature and then determines the numerical value of the feature data parameter to adjust the direction.

7. The intelligent adaptive calibration method for a sensor according to claim 6, wherein: In S105, a preset mathematical model is used to predict the matching options of the sensor, and a matching request corresponding to the actual output value of the specified matching option is detected, including: Establish a mathematical model to describe the error characteristics of the sensor, and then use the mathematical model to predict the parameter output value of the sensor; Obtain the parameter output values ​​of the sensor predicted by the mathematical model, and arrange the parameter output values ​​according to the preset parameter arrangement method to form matching options; Detecting the spectrum state of the sensor in the time series and the actual output value path of the wavelet transform state data and the integrity of the data after adjustment; If it is checked that the actual output value path and data of the sensor's spectrum state and wavelet transform state data in the time series are complete, then a matching option is specified to establish a corresponding matching request.

8. The intelligent adaptive calibration method for a sensor according to claim 7, wherein: In S106, if the specified matching option does not correspond to the matching request, a feedback signal is output to the adjustment unit to determine whether the actual output value path and data of the spectral state and wavelet transform state data of the sensor in the time series after adjustment are incomplete; If the actual output numerical path and data of the spectrum state and wavelet transform state data of the adjusted sensor in the time series are incomplete, the corresponding matching request will continue to be requested within the cycle time until the matching request corresponding to the specified matching option is completed.

9. Intelligent adaptive calibration system for sensors, characterized in that, The intelligent adaptive calibration method according to any one of claims 1 to 8 is applied to the system comprising: A receiving and identifying module is configured to receive a calibration request from a sensor, input the spectral state and wavelet transform state data of the sensor in a time series into a preset data classification model, and obtain a classification result based on the spectral state and wavelet transform state data output by the sensor in the time series; The data analysis module is used to classify the sensor's spectrum state and wavelet transform state data output in the time series, place the classification results in the adjustment unit to complete data comparison, and obtain the deviation error of the benchmark feature data; A correction acquisition module is used to place the deviation error of the reference characteristic data into the adjustment unit to adjust the sensor deviation and obtain the deviation mark; a model matching module, configured to identify the adjusted spectral state and wavelet transform state data of the sensor in a time series, place the data in a preset mathematical model, and predict matching options of the sensor using the preset mathematical model; A prediction comparison module, configured to determine whether a matching option corresponds to the matching request, and if so, to compare the actual output value of the sensor in the matching request with the predicted output value of the sensor; The compensation module is used to compensate and correct the sensor data again according to the current actual output value.

10. The intelligent adaptive calibration system for sensors according to claim 9, characterized in that: The model matching module includes: A model generation unit is used to obtain the system's preset functional parameters and generate a preset mathematical model based on the preset functional parameters; An output matching unit, configured to predict the matching options of the sensor using a preset mathematical model and detect whether a matching request corresponding to an actual output value exists for the specified matching option; The request acquisition unit is configured to determine whether a matching request corresponds to a specified matching option, and compare the actual output value of the sensor in the matching request with the predicted output value of the sensor.

Images