Intelligent calibration method, system and equipment for ambient air monitoring sensor

Abstract

The invention discloses an intelligent calibration method, system and equipment for an ambient air monitoring sensor. The method comprises the following steps: acquiring data of the ambient air monitoring sensor and data of a standard detector in real time and preprocessing the data; performing correlation analysis on the preprocessed data to obtain data in a preset range; performing spatial interpolation calculation and prediction on the data in the preset range to obtain an interpolation estimation value and an air concentration field; re-fitting the air concentration field for the interpolation estimation value to obtain distribution and hour mean value data of the pollutants in the environmental space; performing multi-order linear regression analysis on the obtained hourly mean value data and the hourly mean value data of the sensor to obtain a calibration coefficient; and carrying out calibration calculation by adopting the calibration model to obtain sensor calibration data consistent with the standard detector. The system comprises a module corresponding to the method, and the device comprises a first detection unit, a second detection unit and a processing unit, so that the calibration precision and the calibration efficiency of the sensor can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental monitoring, and in particular to an intelligent calibration method, system and equipment for an ambient air monitoring sensor. Background Art

[0002] Grid-based monitoring of ambient air quality is of great significance in areas such as air quality management and carbon emission control. At present, ambient air sensors are widely used in monitoring stations and grid-based monitoring systems to obtain real-time data on gas concentrations in the environment. Grid-based monitoring divides the target area into multiple small monitoring grids and sets monitoring points in each grid to monitor the concentration of specific pollutants in real time. This method can provide data with high temporal and spatial resolution, which helps to more accurately analyze pollution sources, pollutant diffusion trends, and environmental changes in the region. However, sensors are susceptible to drift, aging, and changes in environmental conditions during long-term use, resulting in inaccurate measurement data, so regular calibration is required to ensure monitoring accuracy.

[0003] Traditional sensor calibration methods are mainly divided into two categories: laboratory calibration and field calibration. Laboratory calibration requires the sensor to be disassembled from the site and transported to the laboratory, where it is calibrated using standard gases under controlled conditions. This method can provide higher calibration accuracy, but the operation is complex and time-consuming, and the transportation and reinstallation of the sensor increase the operation and maintenance costs. In addition, the frequency of laboratory calibration is low, making it difficult to respond to changes in sensor performance in real time, and it cannot meet the needs of large-scale monitoring of sensors. Field calibration is performed directly at the sensor installation location, and calibration is performed using standard gas cylinders and related equipment, avoiding the trouble of disassembling the sensor. Although it improves operation and maintenance efficiency, this method is costly, the equipment is bulky, and energy-intensive, especially in remote areas or places with unstable power supply, where maintenance is difficult.

[0004] Therefore, no matter which method is used, traditional calibration methods are difficult to perform frequently and cannot meet the calibration needs of large-scale deployment of sensors. They have limitations in operation and maintenance efficiency, real-time performance, and cost control. Summary of the invention

[0005] In order to solve some or all of the technical problems existing in the above-mentioned prior art, the present invention provides an intelligent calibration method, system and equipment for an ambient air monitoring sensor, which can improve the calibration accuracy and calibration efficiency of a grid detection sensor.

[0006] The technical solution of the present invention is as follows:

[0007] In a first aspect, the present invention provides an intelligent calibration method for an ambient air monitoring sensor, comprising:

[0008] Acquire sensor data and standard detector data of ambient air monitoring in real time, and pre-process the acquired sensor data and standard detector data;

[0009] Perform correlation analysis on the preprocessed sensor data and standard detector data to obtain data with correlation within a preset range;

[0010] Perform spatial interpolation calculation and prediction on the hourly mean data with correlation within the preset range to obtain the interpolation estimate of the sensor data and the ambient air concentration field at the location monitored by the sensor;

[0011] Refitting the ambient air concentration field at the location monitored by the sensor according to the obtained interpolation estimate value to obtain the distribution of pollutants in the ambient space at the location monitored by the sensor and the hourly average data after refitting the ambient air concentration field at the location monitored by the sensor;

[0012] Perform multi-order linear regression analysis on the hourly average data of the ambient air concentration field at the location monitored by the sensor after refitting and the hourly average data of the sensor in the ambient air monitoring acquired in real time to obtain the calibration coefficient;

[0013] The obtained calibration coefficient is calibrated and calculated using the calibration model to obtain calibration data that is consistent with the sensor data and the standard detector.

[0014] Furthermore, in the above-mentioned intelligent calibration method for ambient air monitoring sensors, the preprocessing includes filtering and denoising the acquired sensor data and standard detector data.

[0015] Furthermore, in the above-mentioned intelligent calibration method for ambient air monitoring sensors, filtering the acquired sensor data and standard detector data includes:

[0016] Calculate the mean and standard deviation of the window data, and calculate the degree of deviation from the mean value in the window based on the calculated mean and standard deviation;

[0017] Determining whether the calculated deviation from the mean within the window exceeds a preset deviation of a set value;

[0018] If the preset deviation is exceeded, the calculated data point will be considered as an outlier and removed;

[0019] The mean is calculated by the following formula:

[0020]

[0021] In the above formula, represents the average value, w represents the window length, t represents the number of data in the window, i represents the starting value of the data in the window, and x iRepresents the i-th number in the window data;

[0022] The standard deviation is calculated by the following formula:

[0023]

[0024] In the above formula, σ t represents standard deviation;

[0025] The degree of deviation from the mean within the window is calculated using the following formula:

[0026]

[0027] In the above formula, Z t Indicates the degree of deviation from the mean value within the window, x t Represents any data in the window.

[0028] Furthermore, in the above-mentioned intelligent calibration method for ambient air monitoring sensors, the preset deviation includes any multiple of 0.5-5 of the mean deviation of the window data exceeded by any data in the window.

[0029] Furthermore, in the above-mentioned intelligent calibration method for ambient air monitoring sensors, the preset range includes data whose correlation between sensor data and standard detector data is between 0.7 and 1.0.

[0030] Furthermore, in the above-mentioned intelligent calibration method for ambient air monitoring sensors, performing spatial interpolation calculation on hourly mean data with correlation within a preset range includes:

[0031] Calculate the semivariance of the standard detector data;

[0032] The variogram model was used to fit the semivariance data, and the variogram model was used to establish the Kriging model;

[0033] Calculate interpolated estimates and error variances of target locations;

[0034] The variogram model includes:

[0035]

[0036] In the above formula, γ(h) represents the variance function model, C 0 Indicates the sill value, C 1 represents the degree of variability, and a represents the range;

[0037] Among them, the Kriging model includes:

[0038]

[0039] In the above formula, λ1 ,λ 2 , …, λ n , represents the weight, C(x,x i ) represents the estimated position (x, y) and each known point (x i ,y i ), μ represents the Lagrange multiplier;

[0040] The interpolated estimated value of the target position is calculated using the following formula:

[0041]

[0042] In the above formula, (x * ,y * ) represents the target position coordinates of the standard detector in the target ambient air monitoring, λ i Represents the calculated weight, which represents the known position relative to the target position (x * ,y * )’s contribution, Indicates the target position coordinates (x * ,y * ), Z i Represents the observed value of a known point;

[0043] Solve the above Kriging model and get the weight λ 1 ,λ 2 , …, λ n , substitute the obtained weight into the interpolation estimate of the target position, and the estimated value of the target position is the standard value of the gridded sensor.

[0044] Furthermore, in the above-mentioned intelligent calibration method for the ambient air monitoring sensor, the calibration model is shown in the following formula:

[0045] R = a 0 +a 1 w+a 2 w 2 +a 3 w 3 +……a n w n ;

[0046] Among them, a 0 , a 1 , a 2 , ..., a n represents the coefficients given by multi-order linear regression, w represents the instantaneous original value of the gridded sensor, R represents the instantaneous calibrated value of the gridded sensor, x, x 2 , x 3, ..., x n Represents standard detector data in ambient air monitoring environment.

[0047] In a second aspect, the present invention provides an intelligent calibration system for an ambient air monitoring sensor, comprising:

[0048] A preprocessing module, the preprocessing module is used to obtain sensor data and standard detector data of ambient air monitoring in real time, and preprocess the obtained sensor data and standard detector data;

[0049] A correlation analysis module, which is used to perform correlation analysis on the preprocessed sensor data and the standard detector data to obtain data with correlation within a preset range;

[0050] A concentration field simulation module, which is used to perform spatial interpolation calculation and prediction on hourly mean data with correlation within a preset range to obtain interpolation estimation values ​​of sensor data and ambient air concentration field at the location monitored by the sensor;

[0051] A comprehensive processing module, the comprehensive processing module is used to refit the ambient air concentration field of the location monitored by the sensor according to the obtained interpolation estimation value, to obtain the distribution of pollutants in the ambient space of the location monitored by the sensor and the hourly average data after refitting the ambient air concentration field of the location monitored by the sensor;

[0052] A linear regression analysis module, wherein the linear regression analysis module is used to perform a multi-order linear regression analysis on the hourly average data obtained after refitting the ambient air concentration field at the location monitored by the sensor and the hourly average data of the sensor in the ambient air monitoring acquired in real time to obtain a calibration coefficient;

[0053] The calibration module is used to perform calibration calculation on the obtained calibration coefficient using the calibration model to obtain calibration data that is consistent with the sensor data and the standard detector.

[0054] In a third aspect, the present invention provides an intelligent calibration device for an ambient air monitoring sensor, comprising:

[0055] A first detection unit, wherein the number of the first detection units is set to be multiple, and the multiple first detection units are evenly spaced and arranged in a grid divided by the ambient air monitoring environment, and are used to monitor the air quality in the ambient air;

[0056] A second detection unit, wherein the number of the second detection units is set to be multiple, and the multiple second detection units are evenly spaced and arranged in a grid divided by the ambient air monitoring environment, and are used to monitor air data in the ambient air;

[0057] A processing unit, the processing unit is wirelessly connected to the first detection unit and the second detection unit, and is equipped with a data acquisition module, a data processing module, a data storage module, a data analysis module and a display module, wherein the data acquisition module is used to receive the data in the ambient air collected by the first detection unit and the second detection unit; the data processing module is used to pre-process the collected data, and after the pre-processing, perform a correlation analysis on the ambient air data obtained by the first detection unit and the ambient air data obtained by the second detection unit to obtain data with a correlation within a preset range, and at the same time perform spatial interpolation calculation and prediction on the data obtained within the preset range to obtain an interpolation estimation value and air concentration of the ambient space where the second detection unit and the first detection unit are located field, and refit the ambient air concentration field in combination with the interpolation estimate value to obtain the distribution of pollutants in the ambient space and the hourly mean data after refitting the ambient air concentration field, and at the same time perform multi-order linear regression analysis in combination with the hourly mean data of the second detection unit to obtain the calibration coefficient, and use the stored, established or called calibration model to perform calibration calculation to obtain the calibration data of the second detection unit data consistent with the standard first detection unit; the data storage module is used to store various data processed by the data processing module; the data analysis module is used to analyze the data correlation performed by the data processing module, and at the same time calculate the spatial interpolation and calibration data performed by the data processing module; the display module is used to display the results processed by the data analysis module and the data processing module.

[0058] The main advantages of the technical solution of the present invention are as follows:

[0059] The intelligent calibration method for an ambient air monitoring sensor of the present invention filters and performs correlation analysis on the data of standardized sensors and sensors, performs spatial interpolation and prediction on the data in a period with high correlation, and finally obtains a calibration coefficient through linear regression analysis to calibrate the sensor data so that the calibrated sensor data is consistent with the measurement results of a standard monitor, with high calibration accuracy, strong real-time performance, low calibration cost and high calibration efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The drawings described herein are used to provide a further understanding of the embodiments of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0061] Figure 1 A schematic diagram of a flow chart of an intelligent calibration method for an ambient air monitoring sensor according to an embodiment of the present invention;

[0062] Figure 2A schematic diagram of the structure of an intelligent calibration device for an ambient air monitoring sensor provided in one embodiment of the present invention.

[0063] Description of reference numerals:

[0064] 1. First detection unit; 2. Second detection unit; 3. Processing unit. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0066] The following is combined with Figure 1-2 , describes in detail the technical solution provided by the embodiments of the present invention.

[0067] First, as attached Figure 1 As shown, an embodiment of the present invention provides an intelligent calibration method for an ambient air monitoring sensor, the method comprising:

[0068] Acquire sensor data and standard detector data of environmental air monitoring in real time, and preprocess the acquired sensor data and standard detector data; perform correlation analysis on the preprocessed sensor data and standard detector data to obtain data with correlation within a preset range; perform spatial interpolation calculation and prediction on the hourly mean data with correlation within the preset range to obtain interpolation estimation values ​​of the sensor data and the environmental air concentration field of the sensor monitoring location; refit the environmental air concentration field of the sensor monitoring location according to the obtained interpolation estimation values ​​to obtain the distribution of pollutants in the environmental space of the sensor monitoring location and the hourly mean data after refitting the environmental air concentration field of the sensor monitoring location; perform multi-order linear regression analysis on the hourly mean data after refitting the environmental air concentration field of the sensor monitoring location and the hourly mean data of the sensor in the environmental air monitoring acquired in real time to obtain calibration coefficients; perform calibration calculation on the obtained calibration coefficients using a calibration model to obtain calibration data consistent with the sensor data and the standard detector.

[0069] Specifically, the intelligent calibration method for ambient air monitoring sensors provided by the embodiments of the present invention is how to calibrate the sensor, and its calibration principle includes:

[0070] The sensor data for ambient air monitoring is calibrated based on the data from the standard detector for ambient air monitoring. During the calibration process, data filtering and correlation analysis, spatial interpolation calculation, and prediction and linear regression analysis are performed on the data from standardized sensors and sensors. This makes the sensor data consistent with the measurement results of the standard monitor, thereby improving the accuracy of the data monitored by the sensor, thereby improving the accuracy of the sensor calibration, and improving the accuracy of the monitoring data. Remote and automated ambient air sensor calibration can be achieved, reducing the need for on-site operation and manual intervention, and greatly improving calibration efficiency.

[0071] Therefore, the intelligent calibration method for ambient air monitoring sensors of the present invention can improve the calibration accuracy and calibration efficiency of grid detection sensors.

[0072] In some optional implementations of the present embodiment, in order to make the sensor calibration more accurate and the acquired data more reference-oriented, the sensor data and standard detector data of ambient air monitoring are acquired in real time, and before preprocessing them, the area of ​​the monitored ambient air is divided into grids, and ambient air detection equipment such as sensors and standard detectors are set in the divided grids.

[0073] Specifically, in order to make the data obtained from the detectors referenceable and the calibration of the grid sensors accurate, the selected detectors should have high precision and stability, low drift and excellent linear response capabilities to ensure accurate and reliable benchmark data. It is preferred that the selected standard monitors have certified calibration certificates to ensure that their data conform to national or international standards.

[0074] Secondly, in order to make the calibration results of grid sensors using standard detectors more accurate, it is necessary to determine the deployment location of the standard monitor. When selecting the location, it is necessary to consider the actual situation of the grid monitoring area, ensure that the deployment can cover the key points of the entire monitoring area, and the deployment location should be able to reflect the environmental changes in the area and obtain representative monitoring data.

[0075] Finally, since drift and aging are inevitable in standard detectors during long-term use, regular calibration and maintenance of standard monitors is necessary. Regular maintenance and recalibration of sensors can cope with possible drift and aging and maintain the high accuracy and reliability of sensors.

[0076] Therefore, by reasonably arranging a number of standard detectors in the environmental space where the monitored ambient air is located and performing regular inspections and maintenance on them, the data monitored by the arranged standard detectors can be made valid data, ensuring the accuracy of sensor calibration and providing a basis and basis for subsequent monitoring and calibration processes.

[0077] In some optional implementations of this embodiment, in order to reduce the cost of ambient air monitoring, a small number of high-precision standard monitors can be reasonably deployed in the target area of ​​ambient air monitoring to provide high-quality benchmark data for comparison and calibration with other sensors. At the same time, the deployed high-precision standard monitors can be deployed in a fixed or mobile manner, depending on the size of the target area of ​​the monitored air and the size of the grid division, to ensure the correlation between the high-precision standard monitors and the accuracy of spatial interpolation.

[0078] Specifically, in order to make the data obtained by the sensor representative of ambient air monitoring, it is necessary to reasonably deploy sensors in the target area of ​​ambient air monitoring in order to obtain representative, real and reliable spatial data. Therefore, when deploying sensors, it is necessary to divide the target area of ​​the monitored ambient air into uniform grids according to the size of the spatial resolution of the target area. Specifically, the size of each grid unit depends on the actual needs and data accuracy requirements, and sensors suitable for target monitoring are selected for deployment. For example, each grid is divided into 1000x1000 meters.

[0079] It should be noted that when selecting a sensor, its range, sensitivity, linearity and long-term stability need to be considered to match the size of the divided grid and the needs of ambient air monitoring.

[0080] Specifically, in the intelligent calibration method of the ambient air monitoring sensor of the present invention, preprocessing includes filtering and denoising the acquired sensor data and standard detector data to eliminate data anomalies caused by environmental changes or accidental human interference, such as removing outliers and outliers caused by environmental interference, so as to ensure the authenticity and representativeness of the data and improve the calibration accuracy, thereby making the data used for calibration more accurate and improving the accuracy of the calibration results.

[0081] Specifically, in this embodiment, the acquired sensor data and standard detector data are filtered to identify abnormal data through the statistical characteristics of local data, and the identified abnormal data is removed as screening data. The above data filtering includes:

[0082] Calculate the mean and standard deviation of the window data, and calculate the degree of deviation from the mean in the window based on the calculated mean and standard deviation; determine whether the calculated degree of deviation from the mean in the window exceeds the preset deviation of the set value; if it exceeds the preset deviation, the calculated data point is regarded as an outlier and removed; wherein:

[0083] The mean is calculated using the following formula:

[0084]

[0085] In the above formula, represents the average value, w represents the window length, t represents the number of data in the window, i represents the starting value of the data in the window, and x i Represents the i-th number in the window data;

[0086] The standard deviation is calculated using the following formula:

[0087]

[0088] In the above formula, σ t represents standard deviation;

[0089] The degree of deviation from the mean within the window is calculated using the following formula:

[0090]

[0091] In the above formula, Z t Indicates the degree of deviation from the mean value within the window, x t Represents any data in the window.

[0092] Therefore, by calculating the mean, the standard deviation and the degree of deviation to determine abnormal data, the retained data is made more accurate, thereby making the data used for calibration more precise and the calibration result more accurate.

[0093] In some optional implementations of this embodiment, the preset deviation may include any multiple of 0.5-5 by which any data in the window exceeds the mean deviation of the window data; preferably, the preset deviation value is set to 2-4 times, most preferably 3 times, of any data in the window exceeding the mean deviation of the window data.

[0094] Such a setting makes it possible to retain an appropriate amount of data for calculation and analysis during the data filtering process, thereby avoiding the situation where the calculation and analysis results are unrepresentative due to too much filtered data, or the calculation and analysis results are too deviated and unreferenceable due to too little screening.

[0095] In some optional implementations of this embodiment, the preset range includes data having a correlation between sensor data and standard detector data between 0.7 and 1.0.

[0096] Specifically, when performing correlation analysis, it is necessary to calculate the correlation between the standard monitor data and the sensor data based on the filtered time series data of the standard monitor and the sensor to help identify and calibrate the sensor error.

[0097] Therefore, during the calibration process, a correlation analysis is performed on the time series data between the sensor and the standard monitor to determine their data consistency and calibration requirements; the correlation analysis includes correlation calculation and statistics on the filtered data to ensure that the interpolation and calibration results are highly reliable and accurate.

[0098] Specifically, in the intelligent calibration method for an ambient air monitoring sensor of the present invention, performing spatial interpolation calculation on hourly mean data with correlation within a preset range includes:

[0099] Calculate the semivariance of the standard detector data; use the variogram model to fit the semivariance data, and use the variogram model to establish the Colibri model; calculate the interpolation estimate and error variance of the target location;

[0100] In some optional implementations of this embodiment, the semivariance is calculated using the following formula:

[0101]

[0102] In the above formula, γ(h) represents the semi-variance of the standard detector data, h represents the distance between the i-th standard detector and the j-th standard detector in the ambient air detection, i and j represent the standard detectors in the ambient air detection, and z i and z j They represent the observed values ​​when the distance between the i-th standard detector and the j-th standard detector in the ambient air detection is taken as the interval h, Var(z i -z j ) represents the variance of the difference in observed values ​​between the i-th standard detector and the j-th standard detection in ambient air detection;

[0103] The variogram model includes:

[0104]

[0105] In the above formula, γ(h) represents the variance function model, C 0 Indicates the sill value, C 1 represents the degree of variability, and a represents the range;

[0106] Select the three values ​​that minimize the objective function (i.e., the three values ​​with the smallest semivariance) to fit the variogram model and obtain C 0 , C 1 、a.

[0107] According to C 0 , C 1 , a and the variogram model can obtain the value of the Kriging model from any point to the location of the standard detector;

[0108]

[0109] Where C(h) represents the value of the Kriging model from any point to the location of the standard detector.

[0110] Among them, the Kriging model includes:

[0111]

[0112] In the above formula, λ 1 ,λ 2 , …, λ n , represents the weight, C(x,x i ) represents the estimated position (x, y) and each known point (x i ,y i ), μ represents the Lagrange multiplier; through the Kriging model, we can first use the standard value to solve the parameter value of the Kriging model and γ(x 1 ,x 1 ) and then use the Kriging model to solve the model value C(x,x) at any location. 1 ) and finally solve the Kriging model to obtain the above weights. Specifically, in the above Kriging model, the matrix The parameters in the formula are The matrix in the above Kriging model is calculated. The parameters in the formula are Calculated.

[0113] The interpolated estimated value of the target position is calculated using the following formula:

[0114]

[0115] In the above formula, (x * ,y * ) represents the target position coordinates of the standard detector in the target ambient air monitoring, λ i Represents the calculated weight, which represents the known position relative to the target position (x * ,y * )’s contribution, Indicates the target position coordinates (x * ,y * ), Z i Represents the observed value of a known point;

[0116] Solve the above Kriging model and get the weight λ 1 ,λ 2 , …, λ n , substitute the obtained weight into the interpolation estimate of the target position, and the estimated value of the target position is the standard value of the gridded sensor.

[0117] Specifically, the weight λ calculated above is i Calculated by the following method:

[0118] Specifically, the weight λ i It is determined by the covariance matrix and the variogram.

[0119] First, we need to calculate the covariance between the known points.

[0120] Assume that there are n known standard monitors, and the positions of the n standard monitors are (x 1 ,y 1 ), (x 2 ,y 2 ), (x 3 ,y 3 )......(x n ,y n ), the observation value corresponding to each position is Z 1 ,Z 2 ,Z 3 ......Z n ;

[0121] Then calculate the spatial distance between each pair of known positions

[0122] The covariance between each pair of locations is then calculated using a selected variogram (e.g., spherical, exponential, or Gaussian).

[0123] In some optional implementations of this embodiment, the error variance is calculated using the following formula:

[0124]

[0125] In the above formula, σ 2 (x * ,y * ) represents the error variance, x * and *The variables representing the error variance, Cov((x i ,y i ),(x j ,y j )) represents the covariance between point pairs, (x i ,y i ),(x j ,y j ) represent the covariance variables, x i ,y i , x j and j Represent variables respectively.

[0126] Therefore, in the present invention, by using an interpolation method to estimate the sensor data in the grid area and predicting possible sensor measurements, the data can be supplemented and expanded in space, thereby providing a more comprehensive data basis, which can further improve the accuracy of the calibrated results. At the same time, an interpolation algorithm is used to interpolate the spatial data of the correlation between the sensor and the standard monitor to ensure that the calibration results can accurately reflect the monitoring conditions of different spatial locations. It should also be noted that the above-mentioned interpolation algorithms include but are not limited to nearest neighbor interpolation, bilinear interpolation, cubic interpolation, Lagrange interpolation, Newton interpolation, etc.

[0127] It should be noted that after using the spatial interpolation method to generate the gas concentration field, it is also necessary to consider the further impact of meteorological factors on the gas concentration. Meteorological conditions such as temperature, humidity, wind speed, and wind direction have a significant impact on atmospheric diffusion, which will lead to nonlinear changes in gas concentration in space. By acquiring meteorological data such as the above-mentioned wind speed and wind direction in real time and combining the above-mentioned interpolation data, the concentration field can be refitted and the gas concentration at each point in space can be dynamically adjusted, which can further reflect the actual distribution of pollutants in space.

[0128] Specifically, a multi-order linear regression is performed based on the hourly average data after refitting and the hourly average data of the sensor, and a calibration coefficient is obtained through linear regression analysis to adjust the measurement data of the sensor.

[0129] In the intelligent calibration method for the ambient air monitoring sensor of the present invention, the calibration model is shown in the following formula:

[0130] Specifically, based on the standard value Z of the hourly mean data after refitting over a period of time 1 ,Z 2 ,Z 3 ......Z n , and the hourly average data W of the gridded sensors 1 ,W 2 ,W 3 ......Wn ,Perform multi-order linear regression.,Through linear regression analysis, the calibration coefficients are obtained,which are used to adjust the measurement data of the gridded,sensor.

[0131] Given the order n of the fit, construct a calibration formula for the polynomial model:

[0132] R = a 0 +a 1 w+a 2 w 2 +a 3 w 3 +……a n w n ;

[0133] Among them, a 0 , a 1 , a 2 , ..., a n represents the coefficients given by multi-order linear regression, w represents the instantaneous original value of the gridded sensor, R represents the instantaneous calibrated value of the gridded sensor, x, x 2 , x 3 , ..., x n Represents standard detector data in ambient air monitoring environment.

[0134] Therefore, through multi-order linear regression, based on the results of interpolation and prediction, the calibration coefficient is obtained and applied to the calibration of sensor data, fully considering the impact of environmental factors on sensor data. By using the calibration formula to calibrate the monitoring data of the sensor in the corresponding time period, the calibrated data is obtained, which can ensure that the sensor data is consistent with the measurement results of the standard monitor, improve the accuracy of the monitoring data, and then improve the accuracy of the sensor calibration.

[0135] In summary, the intelligent calibration method of the ambient air monitoring sensor provided by the embodiment of the present invention can realize remote and automatic calibration of the ambient air sensor through the combination of the cloud platform and the sensor, reduce the need for on-site operation and manual intervention, and greatly improve the calibration efficiency; by using the real-time collected data for multi-order linear regression and calibration, it can be adjusted in time when the sensor drifts or the performance decreases, so as to ensure the accuracy and real-time of the monitoring data; by deploying high-precision standard monitoring instruments and combining spatial interpolation and multi-order linear regression algorithms, data calibration can be performed more accurately, taking into account the changes in space and time, and ensuring the reliability of the calibration results; by deploying sensors based on the characteristics of grid-based layout, the present invention can effectively support ambient air monitoring and calibration in large areas, and also provide accurate data support for air pollution prevention and control and carbon emission management; remote calibration is realized through the cloud platform, and there is no need to frequently disassemble the sensor for laboratory calibration, which reduces the investment in transportation and human resources. At the same time, the configuration of standard gas cylinders and other complex equipment required for on-site calibration is avoided, and the cost of equipment procurement and energy consumption is reduced. Without sacrificing the calibration accuracy, the cost structure is effectively optimized, and the economy and sustainability of the system are greatly improved.

[0136] In general, the intelligent calibration method for ambient air monitoring sensors provided in the embodiments of the present invention not only improves the calibration efficiency and accuracy of the ambient air monitoring sensors, but also enhances the intelligence level of the system, and is applicable to various climate monitoring and air quality management needs.

[0137] In a second aspect, the present invention also provides an intelligent calibration system for an ambient air monitoring sensor, comprising: a preprocessing module, a correlation analysis module, a concentration field simulation module, a comprehensive processing module, a linear regression analysis module and a calibration module, wherein:

[0138] The preprocessing module is used to obtain sensor data and standard detector data of environmental air monitoring in real time, and to preprocess the obtained sensor data and standard detector data; the correlation analysis module is used to perform correlation analysis on the preprocessed sensor data and standard detector data, and obtain data with correlation within a preset range; the concentration field simulation module is used to perform spatial interpolation calculation and prediction on the hourly average data with correlation within a preset range, and obtain the interpolation estimate of the sensor data and the environmental air concentration field of the sensor monitoring location; the comprehensive processing module is used to re-fit the environmental air concentration field of the sensor monitoring location according to the obtained interpolation estimate, and obtain the distribution of pollutants in the environmental space of the sensor monitoring location and the hourly average data after re-fitting the environmental air concentration field of the sensor monitoring location; the linear regression analysis module is used to perform multi-order linear regression analysis on the hourly average data obtained after re-fitting the environmental air concentration field of the sensor monitoring location and the hourly average data of the sensor in the environmental air monitoring acquired in real time, and obtain the calibration coefficient; the calibration module is used to calibrate the obtained calibration coefficient using the calibration model to obtain calibration data consistent with the sensor data and the standard detector.

[0139] Thirdly, Figure 2 As shown, the present invention also provides an intelligent calibration device for an ambient air monitoring sensor, comprising: a first detection unit 1, a second detection unit 2 and a processing unit 3, wherein:

[0140] The number of the first detection unit 1 and the second detection unit 2 is set to be multiple, and the multiple first detection units 1 are evenly spaced in the grid divided by the ambient air monitoring environment to monitor the air quality in the ambient air; the multiple second detection units 2 are evenly spaced in the grid divided by the ambient air monitoring environment to monitor the air data in the ambient air; the processing unit 3 is wirelessly connected to the first detection unit 1 and the second detection unit 2, and the processing unit 3 is equipped with a data acquisition module, a data processing module, a data storage module, a data analysis module and a display module, wherein the data acquisition module is used to receive the data in the ambient air collected by the first detection unit 1 and the second detection unit 2; the data processing module is used to pre-process the collected data, and after pre-processing, perform correlation analysis on the ambient air data obtained by the first detection unit 1 and the ambient air data obtained by the second detection unit 2 to obtain data with correlation within a preset range, and at the same time The obtained data within the preset range is spatially interpolated and predicted to obtain the interpolation estimated value and air concentration field of the environmental space where the second detection unit 2 and the first detection unit 1 are located, and the ambient air concentration field is refitted in combination with the interpolation estimated value to obtain the distribution of pollutants in the environmental space and the hourly average data after refitting the ambient air concentration field. At the same time, a multi-order linear regression analysis is performed in combination with the hourly average data of the second detection unit 2 to obtain the calibration coefficient, and a calibration calculation is performed using the calibration model stored, established or called to obtain calibration data that is consistent with the data of the second detection unit 2 and the standard first detection unit 1; the data storage module is used to store various data processed by the data processing module; the data analysis module is used to analyze the data correlation performed by the data processing module, and at the same time, the spatial interpolation and calibration data performed by the data processing module are calculated; the display module is used to display the results processed by the data analysis module and the data processing module.

[0141] In some optional implementations of the present embodiment, the above-mentioned first detection unit 1 includes a detector, the second detection unit 2 includes a sensor, the processing unit 3 includes a cloud platform, the data acquisition layer in the cloud platform corresponds to the data acquisition module in the processing unit 3 for data acquisition, the data processing layer in the cloud platform corresponds to the data processing module in the acquisition unit for data processing, the data storage layer in the cloud platform corresponds to the data storage module in the processing unit 3 for data storage, the data analysis layer in the cloud platform corresponds to the data analysis module in the processing unit 3 for data analysis, and the visualization and decision support layer in the cloud platform corresponds to the display module in the processing unit 3 for visualization display.

[0142] In summary, the intelligent calibration equipment for environmental air monitoring sensors of the present invention realizes remote and automated calibration of grid-deployed gas sensors through a cloud platform. The cloud platform integrates functions such as data acquisition, filtering, analysis, interpolation, and calibration to ensure the intelligence of the system and the automated calibration process. It also has the advantages of strong real-time performance and high calibration accuracy, reduces operation and maintenance costs during the calibration process, has the characteristics of large-scale application, improves the calibration efficiency and accuracy of environmental monitoring sensors, and enhances the intelligence level of the system, which is suitable for various climate monitoring and air quality management needs.

[0143] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In addition, "front", "back", "left", "right", "upper" and "lower" in this article are all referenced to the placement state shown in the accompanying drawings.

[0144] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent calibration method for an ambient air monitoring sensor, characterized in that: include: Acquire sensor data and standard detector data of ambient air monitoring in real time, and pre-process the acquired sensor data and standard detector data; Perform correlation analysis on the preprocessed sensor data and standard detector data to obtain data with correlation within a preset range; Perform spatial interpolation calculation and prediction on the hourly mean data with correlation within the preset range to obtain the interpolation estimate of the sensor data and the ambient air concentration field at the location monitored by the sensor; Refitting the ambient air concentration field at the location monitored by the sensor according to the obtained interpolation estimate value to obtain the distribution of pollutants in the ambient space at the location monitored by the sensor and the hourly average data after refitting the ambient air concentration field at the location monitored by the sensor; Perform multi-order linear regression analysis on the hourly average data of the ambient air concentration field at the location monitored by the sensor after refitting and the hourly average data of the sensor in the ambient air monitoring acquired in real time to obtain the calibration coefficient; The obtained calibration coefficient is calibrated and calculated using the calibration model to obtain calibration data that is consistent with the sensor data and the standard detector.

2. The intelligent calibration method for ambient air monitoring sensor according to claim 1, characterized in that: The preprocessing includes filtering and denoising the acquired sensor data and standard detector data.

3. The intelligent calibration method for ambient air monitoring sensor according to claim 2, characterized in that: Filtering the acquired sensor data and standard detector data includes: Calculate the mean and standard deviation of the window data, and calculate the degree of deviation from the mean value in the window based on the calculated mean and standard deviation; Determining whether the calculated deviation from the mean within the window exceeds a preset deviation of a set value; If the preset deviation is exceeded, the calculated data point will be considered as an outlier and removed; The mean is calculated by the following formula: In the above formula, represents the average value, w represents the window length, t represents the number of data in the window, i represents the starting value of the data in the window, and x i Represents the i-th number in the window data; The standard deviation is calculated by the following formula: In the above formula, σ t represents standard deviation; The degree of deviation from the mean within the window is calculated using the following formula: In the above formula, Z t Indicates the degree of deviation from the mean value within the window, x t Represents any data in the window.

4. The intelligent calibration method for ambient air monitoring sensor according to claim 3, characterized in that: The preset deviation includes any multiple of 0.5-5 of the mean deviation of the window data exceeded by any data in the window.

5. The intelligent calibration method for ambient air monitoring sensor according to claim 1, characterized in that: The preset range includes data whose correlation between sensor data and standard detector data is between 0.7 and 1.

0.

6. The intelligent calibration method for ambient air monitoring sensor according to claim 1, characterized in that: The spatial interpolation calculation of hourly mean data with correlation within a preset range includes: Calculate the semivariance of the standard detector data; The variogram model was used to fit the semivariance data, and the variogram model was used to establish the Kriging model; Calculate interpolated estimates and error variances of target locations; The variogram model includes: In the above formula, γ(h) represents the variogram model, C0 represents the sill value, C1 represents the variability, and a represents the range; Among them, the Kriging model includes: In the above formula, λ1, λ2, …, λ n , represents the weight, C(x, x i ) represents the estimated position (x, y) and each known point (x i ,y i ), μ represents the Lagrange multiplier; The interpolated estimated value of the target position is calculated using the following formula: In the above formula, (x * ,y * ) represents the target position coordinates of the standard detector in the target ambient air monitoring, λ i Represents the calculated weight, which represents the known position relative to the target position (x * ,y * )’s contribution, Indicates the target position coordinates (x * ,y * ), Z i Represents the observed value of a known point; Solving the above Kriging model, we get the weights λ1, λ2, …, λ n , substitute the obtained weight into the interpolation estimate of the target position, and the estimated value of the target position is the standard value of the gridded sensor.

7. The intelligent calibration method for ambient air monitoring sensors according to claim 1, characterized in that: The calibration model is shown in the following formula: R=a0+a1w+a2w 2 +a3w 3 +……a n w n ; Among them, a0, a1, a2, ..., a n represents the coefficients given by multi-order linear regression, w represents the instantaneous original value of the gridded sensor, R represents the instantaneous calibrated value of the gridded sensor, x, x 2 , x 3 , ..., x n Represents standard detector data in ambient air monitoring environment.

8. An intelligent calibration system for ambient air monitoring sensors, characterized in that: include: A preprocessing module, the preprocessing module is used to obtain sensor data and standard detector data of ambient air monitoring in real time, and preprocess the obtained sensor data and standard detector data; A correlation analysis module, which is used to perform correlation analysis on the preprocessed sensor data and the standard detector data to obtain data with correlation within a preset range; A concentration field simulation module, which is used to perform spatial interpolation calculation and prediction on hourly mean data with correlation within a preset range to obtain interpolation estimation values ​​of sensor data and ambient air concentration field at the location monitored by the sensor; A comprehensive processing module, the comprehensive processing module is used to refit the ambient air concentration field of the location monitored by the sensor according to the obtained interpolation estimation value, to obtain the distribution of pollutants in the ambient space of the location monitored by the sensor and the hourly average data after refitting the ambient air concentration field of the location monitored by the sensor; A linear regression analysis module, wherein the linear regression analysis module is used to perform a multi-order linear regression analysis on the hourly average data obtained after refitting the ambient air concentration field at the location monitored by the sensor and the hourly average data of the sensor in the ambient air monitoring acquired in real time to obtain a calibration coefficient; The calibration module is used to perform calibration calculation on the obtained calibration coefficient using the calibration model to obtain calibration data that is consistent with the sensor data and the standard detector.

9. An intelligent calibration device for an ambient air monitoring sensor, characterized in that: include: A first detection unit, wherein the number of the first detection units is set to be multiple, and the multiple first detection units are evenly spaced and arranged in a grid divided by the ambient air monitoring environment, and are used to monitor the air quality in the ambient air; A second detection unit, wherein the number of the second detection units is set to be multiple, and the multiple second detection units are evenly spaced and arranged in a grid divided by the ambient air monitoring environment, and are used to monitor air data in the ambient air; A processing unit, the processing unit is wirelessly connected to the first detection unit and the second detection unit, and is equipped with a data acquisition module, a data processing module, a data storage module, a data analysis module and a display module, wherein the data acquisition module is used to receive the data in the ambient air collected by the first detection unit and the second detection unit; the data processing module is used to pre-process the collected data, and after the pre-processing, perform a correlation analysis on the ambient air data obtained by the first detection unit and the ambient air data obtained by the second detection unit to obtain data with a correlation within a preset range, and at the same time perform spatial interpolation calculation and prediction on the data obtained within the preset range to obtain an interpolation estimation value and air concentration of the ambient space where the second detection unit and the first detection unit are located field, and refit the ambient air concentration field in combination with the interpolation estimate value to obtain the distribution of pollutants in the ambient space and the hourly mean data after refitting the ambient air concentration field, and at the same time perform multi-order linear regression analysis in combination with the hourly mean data of the second detection unit to obtain the calibration coefficient, and use the stored, established or called calibration model to perform calibration calculation to obtain the calibration data of the second detection unit data consistent with the standard first detection unit; the data storage module is used to store various data processed by the data processing module; the data analysis module is used to analyze the data correlation performed by the data processing module, and at the same time calculate the spatial interpolation and calibration data performed by the data processing module; the display module is used to display the results processed by the data analysis module and the data processing module.

Images