CN111693648A - Internet of things air quality micro sensing data quality checking method and equipment thereof - Google Patents

Internet of things air quality micro sensing data quality checking method and equipment thereof Download PDF

Info

Publication number
CN111693648A
CN111693648A CN202010460673.0A CN202010460673A CN111693648A CN 111693648 A CN111693648 A CN 111693648A CN 202010460673 A CN202010460673 A CN 202010460673A CN 111693648 A CN111693648 A CN 111693648A
Authority
CN
China
Prior art keywords
threshold value
value
low threshold
max
measured values
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202010460673.0A
Other languages
Chinese (zh)
Other versions
CN111693648B (en
Inventor
邓秀明
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Global Global Earth Information Co ltd
Original Assignee
Global Global Earth Information Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Global Global Earth Information Co ltd filed Critical Global Global Earth Information Co ltd
Priority to CN202010460673.0A priority Critical patent/CN111693648B/en
Publication of CN111693648A publication Critical patent/CN111693648A/en
Application granted granted Critical
Publication of CN111693648B publication Critical patent/CN111693648B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/0004—Gaseous mixtures, e.g. polluted air
    • G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
    • G01N33/007—Arrangements to check the analyser
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
    • G01N15/06—Investigating concentration of particle suspensions

Landscapes

  • Chemical & Material Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Pathology (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Immunology (AREA)
  • Physics & Mathematics (AREA)
  • Dispersion Chemistry (AREA)
  • Combustion & Propulsion (AREA)
  • Food Science & Technology (AREA)
  • Medicinal Chemistry (AREA)
  • Investigating Or Analysing Biological Materials (AREA)
  • Automatic Analysis And Handling Materials Therefor (AREA)

Abstract

The invention provides a method and equipment for checking the quality of Internet of things air quality micro sensing data, which mainly utilizes the minimum value of a plurality of measured values to set a preliminary low threshold value yminAnd setting a threshold value y of initial height for the measured value of the maximum valuemaxAnd taking the intercept of a linear regression line obtained according to the measured values as an estimated value
Figure DDA0002510819110000011
Thereafter, the estimated value is reused
Figure DDA0002510819110000012
And said preliminary low threshold value yminObtaining an adjusted low threshold value y'minThe estimated value
Figure DDA0002510819110000013
And the initial height threshold value ymaxObtaining an adjusted high threshold value y'maxAnd then according to the preliminary low threshold value yminAnd the adjusted low threshold value y'minDetermining a low threshold value
Figure DDA0002510819110000014
The initial height threshold value ymaxAnd the adjusted high threshold value y'maxDetermining a high threshold value
Figure DDA0002510819110000015
Thereby, the low threshold value can be determined
Figure DDA0002510819110000016
The high threshold value
Figure DDA0002510819110000017
Checking whether each measuring point is abnormal.

Description

Internet of things air quality micro sensing data quality checking method and equipment thereof
Technical Field
The present invention relates to a method for checking air quality, and more particularly to a method and apparatus for checking the quality of internet-of-things air quality by using a micro-sensor for checking the quality of air.
Background
Under the era trend of the internet of things and the emphasis on air quality, in recent years, a large number of micro sensors are built in various countries to monitor the pollution state of air, especially the particulate matter PM harmful to human body2.5And monitoring of gaseous Volatile Organic Compounds (VOCs).
Before using micro-sensing data, the most important task of the first step is data quality check, which must eliminate the improper data. Otherwise, if inappropriate data is used, the last data produced may result in incorrect results. Because the micro sensing data has a large amount of data characteristics of high time (a few minutes of data) and high space (measuring points are hundreds of meters apart) resolution, an objective and simple method is needed for checking the data quality to meet the use requirement of the data.
The object of the description of the "" air quality micro-sensing data quality check method "", is to provide PM2.5And VOC and other micro-sensing data. The logic of this method assumes that, under long-term averaging, the spatial distribution of data measurements becomes continuous and homogeneous (homogeneous) due to air transport and diffusion, with no unusual values appearing. When the measured value for a long time (more than 6 hours) is too small or too large when compared with the neighboring data, the data is likely to be an instrument abnormality or the measured value is not spatially representative. The reason for this may be that the measuring point is not exactly locatedWhen the measured value is abnormal (the instrument is in a room with too clean air or in a polluted environment for a long time), or the instrument is aged, the instrument is in poor ventilation, or the instrument is in a fault. The checking method is to observe the measured value data of a group of adjacent spaces, and when the measured value of a certain measuring point is too small or too large in the group of data for a long time, or the measured value data of the measuring point is estimated by using the measured value data of other adjacent measuring points, and the long-time deviation between the estimated value and the actual measured value is too large, the measured value of the point is judged to be abnormal.
In the past literature, there are two main methods for quality check of micro-sensing data:
the first method is to sort a group of spatially adjacent measured values into quartiles according to the measured values, and when the measured value at a certain point for a long time is greater than a certain threshold value (maximum quantile +1.5 × (maximum quantile-minimum quantile)) (referred to herein as a high threshold value) or less than a certain threshold value (minimum quantile-1.5 × (maximum quantile-minimum quantile)) (referred to herein as a low threshold value), it is determined that the measured values at the certain point are abnormal (too small or too large for long time representation).
The second method is to use a complex artificial intelligence neural network technology to estimate or predict the measured value data (estimated value) of the measuring points by using the measured value data (actual measured value) of the adjacent measuring points and the data of the previous period, and when the deviation between the estimated value and the actual measured value is too large, it is determined that the measured value data of the measuring points is abnormal.
In practical applications, the first checking method is too severe to check only very high or very low measured values, or some abnormal measured values cannot be checked. The second method is limited to special space-time data due to too complicated procedure (requiring training of a large amount of data), and cannot be popularized in practice in a large amount.
Therefore, in order to solve the above problems in the prior art, a technique for effectively, quickly and simply detecting whether each measurement value is abnormal is needed.
Disclosure of Invention
The invention aims to adopt the advantages of the background technology and avoid the disadvantages thereof, so as to divide the measured values obtained from a plurality of measuring points into a plurality of quantiles to obtain preliminary high and low threshold values, calculate the estimated values of the measured values by adopting a quick and simple statistical regression method, calculate the adjusted high and low threshold values by the estimated values, finally judge the high and low threshold values by utilizing the preliminary high and low threshold values and the adjusted high and low threshold values, and further check whether the measured values are abnormal or not by utilizing the high and low threshold values.
To achieve the above-mentioned object, the present invention provides a method for checking the quality of air quality micro-sensing data of the internet of things, comprising the steps of: taking a circular area with a set radius according to a reference measuring point as a center to obtain a measured value measured by a plurality of measuring points according with a checking quantity in a measuring period; sequencing the measured values from small to large in sequence; setting a preliminary low threshold value y according to the minimum value in the sequence of the plurality of measured valuesmin(ii) a Setting an initial height threshold y for the measured value according to the maximum value in the sequence of the measured valuesmax(ii) a Obtaining a linear regression line according to the measured values, and making an estimation value of the measured values according to the intercept of the linear regression line
Figure BDA0002510819090000021
Using the estimated value
Figure BDA0002510819090000022
Calculating the preliminary low threshold value yminTo obtain an adjusted low threshold value y'min(ii) a According to the adjusted low threshold value y'minAnd the preliminary low threshold value yminDetermining a low threshold value
Figure BDA0002510819090000023
If the adjusted low threshold value y'minLess than said preliminary low threshold value yminThen, the low threshold value is judged
Figure BDA0002510819090000024
Is equal to the adjusted low threshold value y'minIf said adjusted low threshold value y'minGreater than said preliminary low threshold value yminThen, the low threshold value is judged
Figure BDA0002510819090000025
Is equal to the preliminary low threshold value ymin(ii) a Using the estimated value
Figure BDA0002510819090000026
Calculating the preliminary high threshold value ymaxTo obtain an adjusted high threshold value y'max(ii) a According to the adjusted high threshold value y'maxAnd the initial height threshold value ymaxDetermine a high threshold value
Figure BDA0002510819090000027
If the adjusted high threshold value y'maxIs greater than the initial height threshold value ymaxThen, the high threshold value is judged
Figure BDA0002510819090000028
Is equal to the adjusted high threshold value y'maxIf the adjusted high threshold value y'maxLess than the initial height threshold value ymaxThen, the high threshold value is judged
Figure BDA0002510819090000035
Is equal to the initial height threshold value ymax(ii) a And according to the low threshold value
Figure BDA0002510819090000033
And the high threshold value
Figure BDA0002510819090000034
Checking any of the measured values (where any of the measured values is a measured value including any of the measured points within or outside the circular area) when the measured value is less than the low threshold value
Figure BDA0002510819090000036
Or is greater than the high threshold value
Figure BDA0002510819090000037
If so, determining that the measured value is an abnormal value.
Preferably, when the measured values are obtained, the method for checking the quality of the internet-of-things air quality micro sensing data further includes: judging whether the measured values accord with a plurality of checking conditions or not, when the measured values accord with the checking conditions, sequencing the measured values, and when the measured values do not accord with any checking conditions, re-acquiring the measured values measured on the measuring points according with the checking quantity.
Preferably, the checking condition comprises: the actual data number N acquired by the measuring points in a measuring period is more than or equal to the maximum data number N/2 acquired by the measuring points in the measuring period; the measurement period average of the plurality of measured values
Figure BDA0002510819090000038
Or the standard deviation sigma of the time series of the measured values in the measuring periodyNot less than 1.0; or the standard deviation sigma of the time series of the measured values in the measuring periody＞20。
Preferably, the preliminary low threshold value y is setminIn time, the method for checking the quality of the internet-of-things air quality micro sensing data further comprises the following steps: deleting at least one measured value of the maximum value and at least one measured value of the minimum value in the sequence of the measured values; wherein the preliminary low threshold value yminThe initial high threshold value y is set according to the minimum value in the sequence of the measured values after the deleting action is executedmaxThe measured value is set according to the maximum value in the ordering of the plurality of measured values after the deletion action has been performed.
Preferably, the preliminary low threshold value ymin0.5 times the minimum value of the measurement.
Preferably, the initial high threshold value ymaxMultiply the measurement of the maximum value by 1.3.
Preferably, the low threshold value y 'is adjusted'minIs based on the estimated value
Figure BDA0002510819090000031
Calculated as a result.
Preferably, the high threshold value y 'is adjusted'maxIs based on the estimated value
Figure BDA0002510819090000032
Calculated as a result.
Preferably, the measurement period is one of 30 minutes, 1 hour and 2 hours.
Another objective of the present invention is to divide the measured values obtained from multiple measuring points into a plurality of quantiles to obtain a preliminary high-low threshold, and calculate the estimated values of the measured values by a fast and simple statistical regression method, and then calculate an adjusted high-low threshold from the estimated values, and finally determine the high-low threshold by using the preliminary high-low threshold and the adjusted high-low threshold, and further check whether each measured value is abnormal by using the high-low threshold.
To achieve another objective of the above disclosure, the present invention provides an internet of things air quality micro sensing data quality checking device for performing the above-mentioned internet of things air quality micro sensing data quality checking method.
Drawings
FIG. 1 is a schematic diagram of the present invention for obtaining a plurality of measuring points in a circular area;
FIG. 2 is a schematic diagram of a linear regression line according to a plurality of measured values in accordance with the present invention;
FIG. 3 is a schematic view of the present invention for checking whether or not the measured values are abnormal;
FIG. 4 is a flow chart of the steps of the present invention.
Illustration of the drawings:
10: reference measuring point
20: measuring point
CA: circular area
Figure BDA0002510819090000041
Estimated value
ymin: preliminary low threshold
ymax: initial height threshold
Figure BDA0002510819090000042
Low threshold value
Figure BDA0002510819090000043
High threshold value
The process comprises the following steps: S01-S13
Detailed Description
The advantages, features and technical solutions of the present invention will be more readily understood by referring to the exemplary embodiments and the accompanying drawings, which are described in greater detail, and the invention may be implemented in different forms, so it should not be construed that the invention is limited to the embodiments set forth herein, but rather that the embodiments are provided so that this disclosure will fully and completely convey the scope of the invention to those skilled in the art, and that the invention will only be defined by the appended claims.
For a more complete understanding of the present invention, and the advantages achieved, reference is now made to the following descriptions taken in conjunction with the accompanying drawings, in which:
fig. 1 to fig. 3 are schematic diagrams of obtaining a plurality of measurement points in a circular area, a schematic diagram of obtaining a linear regression line according to the plurality of measurement values, and a schematic diagram of checking whether each measurement value is abnormal. As shown in the figure, in order to check each measured value quickly and simply, the present invention takes a circular area CA with a set radius based on a reference measuring point 10 as the center to obtain a plurality of measuring points 20 (e.g. 10 measuring points 20) in accordance with a checking quantity, the selection of the plurality of measuring points 20 is determined according to the measuring point 20 closest to the reference measuring point 10, and in addition, since the measured value data per minute taken by the background art may be influenced by the emission of pollution source, so that the measured value becomes the maximum or minimum value, when the plurality of measuring points 20 are obtained, the present invention obtains a measured value measured by the plurality of measuring points 20 in a measuring period (e.g. 30 minutes, 1 hour or 2 hours, preferably 1 hour), as shown in FIG. 1.
In order to avoid extreme or deviation of the measured values of the measuring points 20 obtained by the above-mentioned operation, a plurality of checking conditions (three checking conditions are provided here) may be provided to filter the measured values, and if the measured values do not meet any of the checking conditions, the measured values that do not meet the checking conditions are discarded, and the measured values measured at the measuring points 20 that meet the checking conditions are re-acquired, for example, if 2 measured values that do not meet the checking conditions are discarded, the measured values at 2 new measuring points 20 are re-acquired in the circular area CA such that the total number of the measured values that are acquired meets the checking number (10) (the discarded measuring points 20 are not included), and if the other measuring points 20 cannot be acquired again in the circular area CA of the previously set setting radius (for example, 10 km), the circular area CA of the set radius (for example, 15 km) is re-enlarged again so that the measuring points 20 corresponding to the number of checks can be re-acquired from within the circular area CA.
Wherein, the checking condition comprises the following steps:
the number N of actual data acquired by the measuring points in a measuring period is more than or equal to the maximum number N/2 of data acquired by the measuring points in the measuring period;
the measurement period average of the plurality of measured values
Figure BDA0002510819090000051
Or the standard deviation sigma of the time series of the measured values in the measuring periodyNot less than 1.0 (not taking the average value too small or not taking the time variation too small); or is
A time series standard deviation σ of the measured values during the measurement periody> 20 (not taking too large a time variation).
When the number of 10 measured values passes through each checking condition, the measured values are sorted in the order from small to large, and then, in order to avoid obtaining too small or too large measured values, at least one of the measured values and at least one of the measured values of the maximum value and the minimum value in the sorting of 10 measured values are deleted, and in this embodiment, 2 measured values of the maximum value and 2 measured values of the minimum value in the sorting of 10 measured values are deleted, so that the remaining measured values are only the measured value of the minimum value in the remaining 6 measured values multiplied by 0.5 to serve as a preliminary low threshold value yminAnd multiplying the largest of the 6 measured values by 1.3 is defined as an initial high threshold ymax。
Thereafter, a linear regression line is obtained by using the measured values as the vertical axis and the distance between the measuring point 20 and the reference measuring point 10 as the horizontal axis, and the intercept of the linear regression line is used as an estimated value of each measured value
Figure BDA0002510819090000061
As shown in fig. 2.
Thus, the estimated value
Figure BDA0002510819090000062
Can be utilized to calculate the preliminary low threshold yminTo obtain an adjusted low threshold value y'minOr is used to calculate the initial step-up threshold ymaxTo obtain an adjusted high threshold value y'maxSpecifically, when the adjustment low threshold value y 'is calculated'minThen, the estimated value can be utilized
Figure BDA0002510819090000063
The calculated adjustment low thresholdValue y'minAnd when calculating the adjusted high threshold value y'maxAt that time, the estimated value can be utilized
Figure BDA0002510819090000064
The calculated adjusted high threshold value y'max。
When the adjusted low threshold value y 'is obtained'minAnd the preliminary low threshold value yminThe time-consuming can use the two to judge a low threshold value
Figure BDA0002510819090000065
Is when the adjusted low threshold value y'minLess than said preliminary low threshold value yminThen, the low threshold value is judged
Figure BDA0002510819090000066
Is equal to the adjusted low threshold value y'minIf said adjusted low threshold value y'minGreater than said preliminary low threshold value yminThen, the low threshold value is judged
Figure BDA0002510819090000067
Is equal to the preliminary low threshold value ymin。
And, similarly, when the adjusted high threshold value y 'is obtained'maxAnd the initial height threshold value ymaxThe two can be used to determine a high threshold value
Figure BDA0002510819090000071
When the high threshold value y 'is adjusted'maxIs greater than the initial height threshold value ymaxThen, the high threshold value is judged
Figure BDA0002510819090000072
Is equal to the adjusted high threshold value y'maxIf the adjusted high threshold value y'maxLess than the initial height threshold value ymaxThen, the high threshold value is judged
Figure BDA0002510819090000073
Is equal to the initial height threshold value ymax。
Thus, when the low threshold value is calculated
Figure BDA0002510819090000074
And the high threshold value
Figure BDA0002510819090000075
Then, the low threshold value can be utilized
Figure BDA0002510819090000076
And the high threshold value
Figure BDA0002510819090000077
Checking the measured value on any measuring point 20 (including any measuring point 20 inside or outside the circular area CA), when the measured value is smaller than the low threshold value
Figure BDA0002510819090000078
Or is greater than the high threshold value
Figure BDA0002510819090000079
Then, the measured value of the measuring point 20 is determined to be an abnormal value, as shown in fig. 4. I.e. when the measured value is smaller than the low threshold value
Figure BDA00025108190900000710
When the measured value is higher than the high threshold value, the measured value is judged to be an abnormal low value in the abnormal values, and when the measured value is higher than the high threshold value
Figure BDA00025108190900000711
And if so, judging that the measured value is an abnormally high value in the abnormal values.
For example, the 10 measured values obtained in the circular area CA are respectively 37, 44, 48, 50, 51, 52, 53, and 53 after sorting, and when deleting two measured values of the maximum value and two measured values of the minimum value in the sorting of the 10 measured values, the measured values of 37, 44 (minimum value) and 53, 53 (maximum value) are respectively deleted, and the remaining measured values such as 48, 50, 51, 52, and 52 are used.
Then, the minimum value (48) of the remaining measured values is multiplied by 0.5 to calculate the preliminary low threshold yminI.e. 48 × 0.5 ═ 24 (y)min)。
Then, the maximum value (52) of the remaining measured values is multiplied by 1.3 to calculate the initial height threshold value ymaxI.e. 52 × 1.3 ≈ 68 (y)max) (rounding off).
In accordance with the present invention, 6 measured values such as 48 (X-0.38, Y-47.8), 50 (X-0.69, Y-49.7), 50 (X-0.67, Y-50.1), 51 (X-0.47, Y-50.8), 52 (X-0.88, Y-51.8), 52 (X-0.92, Y-51.8) (each of the above-mentioned Y values is a value that has not been rounded) are arranged at corresponding positions so that the measured values are vertical axes (Y axes) and the distance between the measuring point 20 and the reference measuring point 10 is horizontal axes (X axes) to obtain a linear regression line, and then the intercept of the linear regression line is used as an estimated value
Figure BDA00025108190900000712
I.e. the estimated value
Figure BDA00025108190900000713
And then based on the estimated value
Figure BDA0002510819090000081
Adjusting low threshold value y'minAnd the adjusted high threshold value y'maxAs follows:
adjusting low threshold value y'min＝47-0.5×(68-24)＝25；
Adjusting high threshold value y'max＝47+0.5*(68-24)＝69。
Calculating the adjusted low threshold value y'minAnd the adjusted high threshold value y'maxThen, the low threshold value can be judged according to the threshold value
Figure BDA0002510819090000082
And the high threshold value
Figure BDA0002510819090000083
As follows:
y′min(＝25)＞ymin(═ 24), so
Figure BDA0002510819090000084
y′max(＝69)＞ymax(68) so
Figure BDA0002510819090000085
Thus, the low threshold value can be determined
Figure BDA0002510819090000086
(24) And the high threshold value
Figure BDA0002510819090000087
(69) And judging whether each measured value is the abnormal value A or not. Assuming that the measured value of a measuring point 20 in the circular area CA is 12 (discarded by each of the examining conditions), the measured value is smaller than the low threshold value
Figure BDA0002510819090000088
(24) And is determined as the abnormal value a (abnormally low value).
Please refer to fig. 4, which is a flowchart illustrating the steps of the creation. As shown in the figure, according to the checking method, the invention can complete the checking method for the air quality micro sensing data quality of the internet of things by the following steps, and the steps comprise:
s01: taking a circular area with a set radius by taking the reference measuring point as a center to obtain a plurality of measuring points according with the checking quantity;
s02: filtering each measured value according to a plurality of checking conditions, if the measured value does not accord with any checking condition, executing the step flow of S13, if all the measured values accord with any checking condition, executing the next step (S03);
s13: discarding the measured value which does not accord with the checking condition, and acquiring the measured value again;
s03: sequencing the measured values according to the sequence from small to large;
s04: deleting at least one measured value of the maximum value and at least one measured value of the minimum value in the sequence of the measured values;
s05: setting a preliminary low threshold value according to a minimum measured value in the sequence of the plurality of measured values;
s06: setting a primary height threshold value according to a maximum value measured value in the sequence of the plurality of measured values;
s07: obtaining a linear regression line according to the measured values, and making estimated values of the measured values according to the intercept of the linear regression line;
s08: calculating a preliminary low threshold value using the estimated value to obtain an adjusted low threshold value;
s09: judging a low threshold value according to the adjusted low threshold value and the preliminary low threshold value, if the adjusted low threshold value is smaller than the preliminary low threshold value, judging that the low threshold value is equal to the adjusted low threshold value, and if the adjusted low threshold value is larger than the preliminary low threshold value, judging that the low threshold value is equal to the preliminary low threshold value;
s10: calculating an initial height threshold value by using the estimated value to obtain an adjusted height threshold value;
s11: judging a high threshold value according to the adjusted high threshold value and the preliminary high threshold value, if the adjusted high threshold value is larger than the preliminary high threshold value, judging that the high threshold value is equal to the adjusted high threshold value, and if the adjusted high threshold value is smaller than the preliminary high threshold value, judging that the high threshold value is equal to the preliminary high threshold value; and
s12: checking any measured value according to the low threshold value and the high threshold value, and judging that the measured value of the measuring point is an abnormal value when the measured value is smaller than the low threshold value or larger than the high threshold value.
The above-mentioned partial steps can be exchanged with each other, but it is not necessary to perform each step in sequence to complete the present invention, for example, the steps can be performed in sequenceFirst, the initial height threshold value y is calculatedmaxThen calculate the initial low threshold yminOr, the adjusted high threshold value y 'may be calculated at first 'maxThen calculates the adjusted low threshold value y'minAnd the like.
According to the technical content disclosed above, the invention can be based on the low threshold value
Figure BDA0002510819090000091
The high threshold value
Figure BDA0002510819090000092
And checking whether each measuring point is abnormal or not, thereby providing a technology for quickly and simply checking each measured value.
The present invention is disclosed in the preferred embodiments, and it is apparent to those skilled in the art that the present invention is not limited thereto.
The present invention is disclosed in the preferred embodiments, and it is apparent to those skilled in the art that the present invention is not limited thereto.

Claims (10)

1. An Internet of things air quality micro sensing data quality checking method is characterized by comprising the following steps:
taking a circular area with a set radius according to a reference measuring point as a center to obtain a measured value measured by a plurality of measuring points according with a checking quantity in a measuring period;
sequencing the measured values from small to large in sequence;
setting a preliminary low threshold value y according to the minimum value in the sequence of the plurality of measured valuesmin；
Setting an initial height threshold y for the measured value according to the maximum value in the sequence of the measured valuesmax；
Obtaining a linear regression line according to the measured values, and calculating the intercept of the linear regression lineAn estimate of the plurality of measurement values
Figure FDA0002510819080000011
Using the estimated value
Figure FDA0002510819080000012
Calculating the preliminary low threshold value yminTo obtain an adjusted low threshold value y'min；
According to the adjusted low threshold value y'minAnd the preliminary low threshold value yminDetermine a low threshold value
Figure FDA0002510819080000013
If the adjusted low threshold value y'minLess than said preliminary low threshold value yminThen, the low threshold value is judged
Figure FDA0002510819080000014
Is equal to the adjusted low threshold value y'minIf said adjusted low threshold value y'minGreater than said preliminary low threshold value yminThen, the low threshold value is judged
Figure FDA0002510819080000015
Is equal to the preliminary low threshold value ymin；
Using the estimated value
Figure FDA0002510819080000016
Calculating the initial height threshold value ymaxTo obtain an adjusted high threshold value y'max；
According to the adjusted high threshold value y'maxAnd the initial height threshold value ymaxDetermine a high threshold value
Figure FDA0002510819080000017
If the adjusted high threshold value y'maxGreater than the initial heightThreshold value ymaxThen, the high threshold value is judged
Figure FDA0002510819080000018
Is equal to the adjusted high threshold value y'maxIf the adjusted high threshold value y'maxLess than the initial height threshold value ymaxThen, the high threshold value is judged
Figure FDA0002510819080000019
Is equal to the initial height threshold value ymax(ii) a And
according to the low threshold value
Figure FDA00025108190800000110
And the high threshold value
Figure FDA00025108190800000111
Checking any of said measured values, when said measured value is less than said low threshold value
Figure FDA00025108190800000113
Or is greater than the high threshold value
Figure FDA00025108190800000112
If so, determining that the measured value is an abnormal value.
2. The method for quality check of internet of things air quality micro sensing data according to claim 1, wherein when the plurality of measured values are obtained, the method for quality check of internet of things air quality micro sensing data further comprises:
judging whether the measured values accord with a plurality of examination conditions, when the measured values accord with the examination conditions, sequencing the measured values, and when the measured values do not accord with any examination conditions, re-acquiring the measured values measured on the measured points which accord with the examination conditions.
3. The internet-of-things air quality micro-sensing data quality checking method according to claim 2, wherein the checking condition includes:
the actual data number N acquired by the measuring points in a measuring period is more than or equal to the maximum data number N/2 acquired by the measuring points in the measuring period;
the measurement period average of the plurality of measured values
Figure FDA0002510819080000021
Or the standard deviation sigma of the time series of the measured values in the measuring periodyNot less than 1.0; or is
A time series standard deviation σ of the measured values during the measurement periody＞20。
4. The method as claimed in claim 1, wherein the preliminary low threshold y is setminIn this case, the method for checking the quality of the internet-of-things air quality miniature sensing data further includes:
deleting at least one measured value of the maximum value and at least one measured value of the minimum value in the sequence of the measured values;
wherein the preliminary low threshold value yminThe initial high threshold value y is set according to the minimum value of the plurality of measured values after the deleting action is executedmaxThe measurement value is set according to the largest value in the ranking of the plurality of measurement values after the deletion action has been performed.
5. The method as claimed in claim 1 or 4, wherein the preliminary low threshold y is set asmin0.5 times the minimum value of the measurement.
6. The Internet of things air quality micro-sensing data quality check of claim 1 or 4Method, characterized in that said initial step-up threshold value ymaxMultiply the measurement of the maximum value by 1.3.
7. The internet of things air quality micro-sensing data quality verification method of claim 1, wherein the adjusting of the low threshold value y'minIs based on the estimated value
Figure FDA0002510819080000022
Calculated as a result.
8. The internet of things air quality micro-sensing data quality verification method of claim 1, wherein the adjusting of the high threshold value y'maxIs based on the estimated value
Figure FDA0002510819080000023
Calculated as a result.
9. The method as claimed in claim 1, wherein the measurement period is one of 30 minutes, 1 hour, and 2 hours.
10. An internet of things air quality micro sensing data quality check device, characterized in that it is used for executing the internet of things air quality micro sensing data quality check method according to any one of claims 1 to 9.
CN202010460673.0A 2020-05-27 2020-05-27 Internet of things air quality micro sensing data quality checking method and equipment thereof Expired - Fee Related CN111693648B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010460673.0A CN111693648B (en) 2020-05-27 2020-05-27 Internet of things air quality micro sensing data quality checking method and equipment thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010460673.0A CN111693648B (en) 2020-05-27 2020-05-27 Internet of things air quality micro sensing data quality checking method and equipment thereof

Publications (2)

Publication Number Publication Date
CN111693648A true CN111693648A (en) 2020-09-22
CN111693648B CN111693648B (en) 2022-12-13

Family

ID=72478604

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010460673.0A Expired - Fee Related CN111693648B (en) 2020-05-27 2020-05-27 Internet of things air quality micro sensing data quality checking method and equipment thereof

Country Status (1)

Country Link
CN (1) CN111693648B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005284405A (en) * 2004-03-26 2005-10-13 Matsushita Electric Works Ltd Abnormality diagnosis apparatus
US20070032706A1 (en) * 2003-08-22 2007-02-08 Apurv Kamath Systems and methods for replacing signal artifacts in a glucose sensor data stream
CN101290611A (en) * 2007-04-20 2008-10-22 中芯国际集成电路制造(上海)有限公司 Data abnormal point detection method and device
US20180225166A1 (en) * 2017-02-03 2018-08-09 Kabushiki Kaisha Toshiba Abnormality detection apparatus, abnormality detection method, and non-transitory computer readable medium
TW201835546A (en) * 2017-03-22 2018-10-01 崑山科技大學 Air quality sensing devices and anti- interfering air quality sensing methods
CN109239265A (en) * 2018-09-11 2019-01-18 清华大学合肥公共安全研究院 Monitoring device fault detection method and device

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070032706A1 (en) * 2003-08-22 2007-02-08 Apurv Kamath Systems and methods for replacing signal artifacts in a glucose sensor data stream
JP2005284405A (en) * 2004-03-26 2005-10-13 Matsushita Electric Works Ltd Abnormality diagnosis apparatus
CN101290611A (en) * 2007-04-20 2008-10-22 中芯国际集成电路制造(上海)有限公司 Data abnormal point detection method and device
US20180225166A1 (en) * 2017-02-03 2018-08-09 Kabushiki Kaisha Toshiba Abnormality detection apparatus, abnormality detection method, and non-transitory computer readable medium
TW201835546A (en) * 2017-03-22 2018-10-01 崑山科技大學 Air quality sensing devices and anti- interfering air quality sensing methods
CN109239265A (en) * 2018-09-11 2019-01-18 清华大学合肥公共安全研究院 Monitoring device fault detection method and device

Also Published As

Publication number Publication date
CN111693648B (en) 2022-12-13

Similar Documents

Publication Publication Date Title
CN109564422B (en) Tool condition monitoring and matching
US10761128B2 (en) Methods and systems for inline parts average testing and latent reliability defect detection
CN106547265B (en) A kind of live reliability estimation method and system of rail traffic electronic control unit
JP2009080612A (en) Method for evaluating distribution, method for producing article, distribution evaluation program, and distribution evaluation system
CN108416817B (en) Automatic residual image correction coefficient obtaining method
JP6941169B2 (en) Quantification and reduction of total measurement uncertainty
CN115861153A (en) Image detection method, computer device and storage medium
CN108023782A (en) A kind of equipment fault early-warning method based on maintenance record
CN107798149B (en) Aircraft maintainability assessment method
CN112668476A (en) Data processing method and device, electronic equipment and storage medium
CN111693648B (en) Internet of things air quality micro sensing data quality checking method and equipment thereof
CN111122775A (en) Pollution concentration monitoring equipment-oriented segmentation data calibration method and system
CN110850358A (en) Electric energy meter comprehensive verification method and system based on stepwise regression algorithm
TWI748486B (en) Quality inspection method for micro-sensing data of air quality of internet of things
CN115112663A (en) Method and apparatus for adjusting condition for determining whether or not a test object is good
CN115408292A (en) Method, device, equipment and medium for evaluating software testing process
CN116646281A (en) Abnormal test structure acquisition method, abnormal test structure verification method and related devices
CN109446481B (en) Method for estimating life distribution parameters of log-normal unit
CN109145502B (en) Weibull unit life distribution parameter estimation method
CN115546131A (en) Quantitative evaluation method for black ash on surface of strip steel and related equipment
CN109375007B (en) Failure evaluation method, device, equipment and storage medium of measurement and control protection device
JP2000321327A (en) Apparatus and method for inspecting semiconductor device
CN115308518B (en) Parameter measurement sequence determination method and system for burn-in circuit
CN114167335B (en) Qualification inspection method and system for newly added detection tool
CN113012120B (en) Gray value validity analysis method and device, electronic equipment and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20221213