CN117574061A - PM2.5 and ozone pollution cooperative prevention and control prediction method and system - Google Patents
PM2.5 and ozone pollution cooperative prevention and control prediction method and system Download PDFInfo
- Publication number
- CN117574061A CN117574061A CN202410057175.XA CN202410057175A CN117574061A CN 117574061 A CN117574061 A CN 117574061A CN 202410057175 A CN202410057175 A CN 202410057175A CN 117574061 A CN117574061 A CN 117574061A
- Authority
- CN
- China
- Prior art keywords
- data
- ozone
- data sequence
- weight
- sequence
- 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
Links
- CBENFWSGALASAD-UHFFFAOYSA-N Ozone Chemical compound [O-][O+]=O CBENFWSGALASAD-UHFFFAOYSA-N 0.000 title claims abstract description 543
- 238000000034 method Methods 0.000 title claims abstract description 50
- 230000002265 prevention Effects 0.000 title claims abstract description 39
- 230000008859 change Effects 0.000 claims description 133
- 230000008569 process Effects 0.000 claims description 17
- YHXISWVBGDMDLQ-UHFFFAOYSA-N moclobemide Chemical compound C1=CC(Cl)=CC=C1C(=O)NCCN1CCOCC1 YHXISWVBGDMDLQ-UHFFFAOYSA-N 0.000 claims description 6
- 230000002195 synergetic effect Effects 0.000 claims 1
- 238000001914 filtration Methods 0.000 abstract description 80
- 238000004458 analytical method Methods 0.000 abstract description 22
- 230000000694 effects Effects 0.000 abstract description 15
- 238000012545 processing Methods 0.000 abstract description 13
- 238000003915 air pollution Methods 0.000 abstract description 10
- 230000000875 corresponding effect Effects 0.000 description 88
- 239000003344 environmental pollutant Substances 0.000 description 15
- 231100000719 pollutant Toxicity 0.000 description 15
- 230000002146 bilateral effect Effects 0.000 description 12
- 238000004364 calculation method Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 5
- 239000002699 waste material Substances 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 238000003672 processing method Methods 0.000 description 3
- 239000000126 substance Substances 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 239000000809 air pollutant Substances 0.000 description 1
- 231100001243 air pollutant Toxicity 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/27—Regression, e.g. linear or logistic regression
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A50/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE in human health protection, e.g. against extreme weather
- Y02A50/20—Air quality improvement or preservation, e.g. vehicle emission control or emission reduction by using catalytic converters
Abstract
The invention relates to the technical field of data filtering analysis, in particular to a PM2.5 and ozone pollution cooperative prevention and control prediction method and system, wherein the size of a first filtering window of an ozone data sequence is determined according to an obtained PM2.5 data sequence and an obtained ozone data sequence and by combining the overall noise influence state of ozone data; analyzing the local noise influence state of the ozone data sequence, determining the noise influence degree of each ozone data in the ozone data sequence, correcting the first filter window size by utilizing the noise influence degree of each ozone data to obtain the optimal filter window size of each ozone data, and further obtaining the ozone data sequence after the filter processing; and determining the prediction results of the PM2.5 and the ozone by using a prediction model according to the PM2.5 data sequence and the filtered ozone data sequence. The invention improves the filtering effect of ozone data and further improves the accuracy of air pollution prevention and control prediction analysis.
Description
Technical Field
The invention relates to the technical field of data filtering analysis, in particular to a prediction method and a prediction system for PM2.5 and ozone pollution cooperative prevention and control.
Background
With the popularization of industry and modern transportation means, the generated waste brings certain pollution to the environment, PM2.5 and ozone in the waste are main atmospheric pollutants in urban areas, wherein the ozone is secondary pollutant, PM2.5 has the properties of primary pollutant and secondary pollutant, and a complex interaction process exists between the two. PM2.5 and ozone in the air are monitored and predicted, and the method plays an important role in environmental management, prevention and control.
When PM2.5 and ozone in the air are monitored and predicted, noise data exists in ozone data monitored by a monitor because the ozone is more easily influenced by factors in the surrounding environment, such as illumination, temperature, air flow and the like, and the authenticity of the ozone data is difficult to guarantee. PM2.5 in air is relatively stable compared to ozone. In order to improve the accuracy of collected ozone data, filtering treatment is carried out on the ozone data, the conventional common non-local bilateral filtering algorithm carries out filtering treatment on the ozone data, the filtering treatment is realized by carrying out local analysis on a window with a fixed size during filtering, the selection of the window size has a certain influence on the final filtering effect, the fixed size filtering window set in the conventional algorithm cannot adapt to different noise influence degrees of different data, the final filtering effect of the ozone data is caused to be poor, the authenticity of the ozone data is still poor, and the accuracy of air pollution prevention and control prediction analysis based on PM2.5 and the filtered ozone data is further caused to be low.
Disclosure of Invention
In order to solve the technical problem that the accuracy of air pollution prevention and control prediction analysis based on PM2.5 and ozone data is low due to the fact that the filtering effect of ozone data corresponding to the filtering algorithm of the conventional fixed-size filtering window is poor, the invention aims to provide a PM2.5 and ozone pollution cooperative prevention and control prediction method and system, and the adopted technical scheme is as follows:
an embodiment of the invention provides a prediction method for PM2.5 and ozone pollution cooperative prevention and control, which comprises the following steps:
acquiring a PM2.5 data sequence and an ozone data sequence; according to the PM2.5 data sequence and the ozone data sequence, determining data change characteristic values corresponding to the PM2.5 data sequence and the ozone data sequence;
correcting a preset initial filter window size according to the data change characteristic value corresponding to the PM2.5 data sequence and the ozone data sequence, the maximum concentration value and the minimum concentration value in the ozone data sequence and the correlation between the PM2.5 data sequence and the ozone data sequence, and determining a first filter window size of the ozone data sequence;
determining the noise influence degree of each ozone data according to the concentration value of each ozone data in the ozone data sequence and the data change characteristic value corresponding to the preset data interval of each ozone data;
Correcting the first filter window size according to preset adjustment parameters, noise influence degree of each ozone data, data change characteristic values corresponding to preset data intervals and data change characteristic values corresponding to ozone data sequences, and determining the optimal filter window size of each ozone data;
according to the optimal filter window size of each ozone data, carrying out filter treatment on the ozone data sequence to obtain a filtered ozone data sequence; and determining the prediction results of the PM2.5 and the ozone by using a prediction model according to the PM2.5 data sequence and the filtered ozone data sequence.
Further, the determining, according to the PM2.5 data sequence and the ozone data sequence, the data change characteristic value corresponding to the PM2.5 data sequence and the ozone data sequence includes:
taking the PM2.5 data sequence or the ozone data sequence as a target gas data sequence, determining standard deviation and mean value of the target gas data sequence, and taking the ratio of the standard deviation to the mean value as a first variation characteristic factor;
calculating the corresponding slope of the adjacent gas data according to each gas data in the target gas data sequence to obtain each slope; determining differences between adjacent slopes according to the slopes to obtain the differences of the slopes, and taking the average value of the differences of all the slopes as a second change characteristic factor;
And combining the first change characteristic factor and the second change characteristic factor to obtain a data change characteristic value corresponding to the target gas data sequence.
Further, the method for determining the first filter window size of the ozone data sequence by correcting the preset initial filter window size according to the data change characteristic value corresponding to the PM2.5 data sequence and the ozone data sequence, the maximum concentration value and the minimum concentration value in the ozone data sequence and the correlation between the PM2.5 data sequence and the ozone data sequence comprises the following steps:
determining the difference between the data change characteristic value corresponding to the PM2.5 data sequence and the data change characteristic value corresponding to the ozone data sequence, and taking the difference between the two data change characteristic values as a first weight;
obtaining a maximum concentration value and a minimum concentration value in an ozone data sequence, calculating a difference value between the maximum concentration value and the minimum concentration value, and taking the difference value between the maximum concentration value and the minimum concentration value as a second weight;
determining a pearson coefficient according to the PM2.5 data sequence and the ozone data sequence, and taking an inverse proportion value of the absolute value of the pearson coefficient as a third weight;
and combining the first weight, the second weight and the third weight to obtain an initial window size weight, upwardly rounding the product of the initial window size weight and a preset initial filter window size, and taking the upwardly rounded value as the first filter window size of the ozone data sequence.
Further, the obtaining the initial window size weight by combining the first weight, the second weight and the third weight includes:
calculating the product of the first weight, the second weight and the third weight, normalizing the product of the first weight, the second weight and the third weight to obtain a normalized value, and taking the normalized value as an initial window size weight.
Further, the determining the noise influence degree of each ozone data according to the concentration value of each ozone data in the ozone data sequence and the data change characteristic value corresponding to the preset data interval of each ozone data includes:
determining the number of concentration values which are the same as the concentration value of the target ozone data in an ozone data sequence by taking any one of the ozone data as the target ozone data, performing inverse operation on the number of concentration values to obtain an inverse proportion value of the number of concentration values, and taking the inverse proportion value of the number of concentration values as a first noise influence factor;
determining preset data intervals of each ozone data, and determining a second noise influence factor according to the data change characteristic value corresponding to the preset data interval of the target ozone data and the data change characteristic values corresponding to the preset data intervals of the previous and next ozone data;
Calculating a concentration value average value corresponding to the ozone data sequence, and taking the difference between the concentration value of the target ozone data and the concentration value average value corresponding to the ozone data sequence as a third noise influence factor;
and fusing the first noise influence factor, the second noise influence factor and the third noise influence factor, and determining the noise influence degree of the target ozone data.
Further, the determining the second noise influence factor according to the data change characteristic value corresponding to the preset data interval of the target ozone data and the data change characteristic values corresponding to the preset data intervals of the previous and the next ozone data includes:
calculating the average value of the data change characteristic values corresponding to the preset data interval of the previous ozone data and the next ozone data of the target ozone data, and taking the difference between the data change characteristic value corresponding to the preset data interval of the target ozone data and the average value of the data change characteristic values as a second noise influence factor.
Further, the preset data interval of the ozone data refers to a preset number of ozone data nearest to the ozone data and a data interval formed by the preset number of ozone data.
Further, the correcting the first filter window size according to the preset adjustment parameter, the noise influence degree of each ozone data, the data change characteristic value corresponding to the preset data interval and the data change characteristic value corresponding to the ozone data sequence, and determining the optimal filter window size of each ozone data includes:
Determining an average value of noise influence degrees of all ozone data; for any one ozone data, determining the ratio between the noise influence degree of the ozone data and the average value of the noise influence degrees of all the ozone data as a fourth weight;
taking the ratio of the data change characteristic value corresponding to the preset data interval of the ozone data to the data change characteristic value corresponding to the ozone data sequence as a fifth weight;
calculating the product of the fourth weight and the fifth weight, normalizing the product of the fourth weight and the fifth weight to obtain a normalized value of the product of the fourth weight and the fifth weight, and determining the normalized value as a final window size weight;
and (3) carrying out upward rounding on the product of the final window size weight, the preset adjusting parameter and the first filter window size, and taking the numerical value obtained after upward rounding as the optimal filter window size of the ozone data.
Further, the determining the prediction result of the PM2.5 and the ozone according to the PM2.5 data sequence and the filtered ozone data sequence by using a prediction model includes:
and (3) taking the PM2.5 data sequence and the filtered ozone data sequence as input data, and inputting the input data into a pre-constructed and trained ARIMA prediction model to obtain a PM2.5 and ozone prediction result.
The embodiment of the invention also provides a PM2.5 and ozone pollution cooperative prevention and control prediction system, which comprises a processor and a memory, wherein the processor is used for processing instructions stored in the memory so as to realize the PM2.5 and ozone pollution cooperative prevention and control prediction method.
The invention has the following beneficial effects:
the invention provides a PM2.5 and ozone pollution cooperative prevention and control prediction method and system, which are mainly applicable to the field of ozone data processing, noise exists in ozone data because collected ozone data is easily influenced by surrounding environment, and in order to improve the accuracy of filtering and denoising of the ozone data and further improve the accuracy of an air pollution prevention and control prediction analysis result, the invention combines the data characteristics of PM2.5 data and ozone data, quantifies the noise influence degree of each ozone data, and adaptively determines the window size of the ozone data when non-local bilateral filtering is used, namely determines the optimal filtering window size of each ozone data. Firstly, when ozone data is changed, PM2.5 data also changes to a certain extent, so that when the first filter window size of the ozone data sequence is analyzed, the numerical accuracy of the first filter window size can be effectively improved by combining the data characteristics of the PM2.5 data; secondly, when the size of the first filtering window is determined, not only the data change characteristic values corresponding to the PM2.5 data sequence and the ozone data sequence are considered, but also the maximum concentration value and the minimum concentration value in the ozone data sequence and the correlation between the PM2.5 data sequence and the ozone data sequence are considered, and the influence degree of noise on the whole ozone data is quantized from multiple angles, so that the reliability degree of the size of the first filtering window is improved; then, the local noise influence in the ozone data sequence is further quantized, namely, the noise influence degree of each ozone data is determined, the noise influence degree of each ozone data is utilized to correct the size of the first filtering window, each ozone data can be caused to have the corresponding self-adaptive optimal filtering window size, and compared with the size of the filtering window which is directly and manually set, the optimal filtering window size of each ozone data determined by the method can adapt to ozone data with different noise influence degrees, the filtering effect of the ozone data is improved, and ozone data with higher accuracy is obtained; based on ozone data and PM2.5 data with higher accuracy, the accuracy of air pollution prevention and control prediction analysis is improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a predictive method for PM2.5 and ozone pollution cooperative prevention and control in an embodiment of the invention;
fig. 2 is a flowchart of determining the noise influence degree of the target ozone data in the embodiment of the invention.
Detailed Description
In order to further describe the technical means and effects adopted by the present invention to achieve the preset purpose, the following detailed description is given below of the specific implementation, structure, features and effects of the technical solution according to the present invention with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
A prediction method and a system embodiment for PM2.5 and ozone pollution cooperative prevention and control:
the application scene aimed by the embodiment of the invention is as follows:
when the predictive analysis is performed based on the ozone data and the PM2.5 data, the collected ozone data is easily influenced by the external environment, so that noise data exists in the ozone data, and the accuracy of the predictive analysis result of the subsequent atmospheric pollutants is influenced by the noise data. Therefore, the collected ozone data needs to be subjected to filtering processing to overcome the influence of noise data in the ozone data. The existing non-local bilateral filtering is generally used for denoising the ozone data, the size of a filtering window influences the accuracy of final filtering of the ozone data, so that characteristic analysis can be performed according to the collected ozone data, the size of the filtering window is self-adaptive, the final filtering effect corresponding to the ozone data is improved, and the accuracy of subsequent analysis and prediction of the data is further improved.
In order to adapt to the window size of ozone data when non-local bilateral filtering is used, and improve the accuracy of ozone data filtering, the embodiment provides a prediction method for PM2.5 and ozone pollution cooperative prevention and control, as shown in fig. 1, which comprises the following steps:
S1, acquiring a PM2.5 data sequence and an ozone data sequence.
It should be noted that, in this embodiment, the denoising process is performed on the ozone data collected in the air, and then the denoised ozone data and the PM2.5 data in the air are further predicted and analyzed, so that prevention and control measures are taken on the pollutants in the air. Therefore, the present embodiment needs to acquire real and accurate PM2.5 data and ozone data, where the PM2.5 data and ozone data may refer to concentration data of PM2.5 and ozone in air.
In this embodiment, PM2.5 acquisition sensor and ozone acquisition sensor that lay through the air monitoring station gather PM2.5 data and ozone data simultaneously, and PM2.5 data and ozone data's collection frequency is once every minute, and the collection period is current a week, constitutes PM2.5 data sequence and ozone data sequence respectively with all data that gather, and every gas data all has its corresponding concentration value, and the amplitude size of gas data can be represented to the concentration value. Wherein, the collection frequency and the collection period are empirical values, and the practitioner can set the collection frequency and the collection period of the data according to specific practical situations, which is not particularly limited herein.
So far, the embodiment obtains the PM2.5 data sequence and the ozone data sequence corresponding to the air monitoring station.
It should be noted that, PM2.5 data in air is relatively stable, and ozone data is easily influenced by surrounding environment, and collected ozone data contains a large amount of noise data, so in order to improve the authenticity of collected ozone data, the subsequent air pollutant prevention and control prediction analysis is convenient, and the embodiment carries out filtering denoising analysis on the collected ozone data, and the specific implementation process comprises the following steps S2 to S5, including:
s2, correcting the preset initial filter window size according to the PM2.5 data sequence and the ozone data sequence, and determining the first filter window size of the ozone data sequence.
It should be noted that ozone and PM2.5 are generally generated in a homologous manner, when ozone data is changed, the data of PM2.5 will also change to a certain extent, that is, the change feature of ozone data has a certain similarity with the change feature of PM2.5 data, so that the first filter window size of the ozone data sequence can be quantified by combining the similarity of the data features between the PM2.5 data sequence and the ozone data sequence, and the data change feature values corresponding to the PM2.5 data sequence and the ozone data sequence can be determined first.
And a first step of determining data change characteristic values corresponding to the PM2.5 data sequence and the ozone data sequence according to the PM2.5 data sequence and the ozone data sequence.
In this embodiment, the data change feature value may represent the fluctuation degree of the data sequence, and for the PM2.5 data sequence and the ozone data sequence, the two sequences are consistent in determining the data change feature value, and any one of the two sequences may be taken as an example, and the PM2.5 data sequence or the ozone data sequence may be taken as the target gas data sequence, so as to determine the data change feature value corresponding to the target gas data sequence, where the specific implementation process includes:
a first sub-step of determining a first variation characteristic factor of the target gas data sequence.
In this embodiment, the standard deviation and the mean of the target gas data sequence are calculated according to each gas data in the target gas data sequence, and the ratio of the standard deviation and the mean is used as the first variation characteristic factor. The calculation process of the standard deviation and the mean value is the prior art, and is not described herein; the first variation characteristic factor may be characterized as a variation coefficient of the target gas data sequence, and the smaller the average value of the target gas data sequence, the larger the standard deviation, the larger the variation coefficient, which indicates that the more the target gas data sequence is dispersed, the larger the fluctuation, the more obvious the data variation characteristic, and the larger the first variation characteristic factor.
It should be noted that, when the variation characteristics of the data sequence are analyzed, the variation coefficient can provide a more meaningful variation description, namely, the first variation characteristic factor, and the determination of the first variation characteristic factor is beneficial to the subsequent determination of the difference of the variation characteristics between the PM2.5 data sequence and the ozone data sequence; meanwhile, the first change characteristic factor is one of indexes for measuring the data change characteristic value corresponding to the target gas data sequence; the first change characteristic factor and the data change characteristic value are in positive correlation, and the larger the first change characteristic factor is, the larger the data change characteristic value is.
A second substep of determining a second varying characteristic factor of the target gas data sequence.
In the embodiment, calculating a slope corresponding to adjacent gas data according to each gas data in the target gas data sequence to obtain each slope; and determining differences between adjacent slopes according to the slopes, obtaining the slope differences, and taking the average value of all the slope differences as a second change characteristic factor. Wherein the difference between adjacent slopes can be determined by calculating the absolute value of the difference between the two slopes.
It should be noted that the second variation characteristic factor may represent the overall variation degree of the target gas data sequence, and the larger all adjacent slope differences are, the larger the data fluctuation degree of the target gas data sequence is, the more obvious the data variation characteristic is; the second change characteristic factor is one of indexes for measuring the data change characteristic value corresponding to the target gas data sequence, the second change characteristic factor and the data change characteristic value are in positive correlation, and the larger the second change characteristic factor is, the larger the data change characteristic value is.
And a third sub-step, combining the first change characteristic factor and the second change characteristic factor to obtain a data change characteristic value corresponding to the target gas data sequence.
In this embodiment, the combination of the first change feature factor and the second change feature factor is achieved by calculating the product of the first change feature factor and the second change feature factor, and the product of the first change feature factor and the second change feature factor is used as the data change feature value corresponding to the target gas data sequence.
As an example, the calculation formula of the data change characteristic value corresponding to the target gas data sequence may be:
the method comprises the steps of carrying out a first treatment on the surface of the Wherein T is a data change characteristic value corresponding to the target gas data sequence,for standard deviation of the target gas data sequence, +.>For the mean value of the target gas data sequence, +.>For the first variation characteristic factor of the target gas data sequence, m is the number of gas data of the target gas data sequence,/I>I is the number of slopes in the target gas data sequence,/>Is the absolute value of the difference between the ith slope and the (i+1) th slope in the target gas data sequence,/v>Is a second varying characteristic factor of the target gas data sequence.
Up to this point, the present embodiment determines the data change characteristic value corresponding to the PM2.5 data sequence and the ozone data sequence.
And secondly, correcting the preset initial filter window size according to the data change characteristic values corresponding to the PM2.5 data sequence and the ozone data sequence, the maximum concentration value and the minimum concentration value in the ozone data sequence and the correlation between the PM2.5 data sequence and the ozone data sequence, and determining the first filter window size of the ozone data sequence.
It should be noted that, in the natural environment, when there is air flow or temperature change in the air, the PM2.5 data and the ozone data will change, and when the air is not affected by the external environment, the ozone change and the PM2.5 change may be relatively close, and the ozone is more easily affected by the external factors and has noise, so that further analysis is needed to determine the first filter window size of the ozone data during filtering by the change characteristics of the ozone and the PM 2.5.
The first filter window size is a window size determined by quantifying the overall noise level of the ozone data sequence according to the overall data change characteristics of the ozone data sequence. In order to improve the filtering effect of the non-local bilateral filtering algorithm, the larger the overall noise degree is, the larger the size of a filtering window of the ozone data sequence is, and at the moment, the sizes of the filtering windows corresponding to all ozone data in the ozone data sequence are consistent.
And a first sub-step of determining a first weight of the initial window size according to the PM2.5 data sequence and the data change characteristic value corresponding to the ozone data sequence.
In this embodiment, a difference between a data change characteristic value corresponding to the PM2.5 data sequence and a data change characteristic value corresponding to the ozone data sequence is determined, and the difference between the two data change characteristic values is taken as the first weight. The difference between the two data change feature values may be determined by the absolute value of the difference between the two data change feature values, although the practitioner may determine the difference between the two data in other ways.
It should be noted that PM2.5 and ozone are pollutant homologies, and the environments where the two data sequences are located are relatively consistent, and the similarity of the data change characteristics corresponding to the two data sequences is higher, so that when the first weight is larger, the degree of influence of noise on the ozone data is larger, and the corresponding first filter window size is also larger.
And a second sub-step of determining a second weight of the initial window size according to the maximum concentration value and the minimum concentration value in the ozone data sequence.
In this embodiment, the concentration value of each ozone data in the ozone data sequence is acquired first, then the maximum concentration value and the minimum concentration value are determined from all the concentration values, then the difference between the maximum concentration value and the minimum concentration value is calculated, and finally the difference between the maximum concentration value and the minimum concentration value is used as the second weight.
It should be noted that, the larger the second weight, the larger the numerical variation range in the ozone data sequence, the larger the noise influence degree, and the larger the corresponding first filter window size. The reason why the larger the numerical variation range is, the larger the noise influence degree is: when the ozone data is not influenced by external factors, the concentration change corresponding to the collected ozone data is kept relatively stable; when ozone data is interfered by external factors, the chemical property of ozone is unstable, the ozone data is more easily interfered by noise, and whether the ozone data is interfered by the external factors or not can be measured through the floating range of the ozone data, namely, the whole noise influence degree of the ozone data is quantized.
And a third sub-step of determining a third weight of the initial window size based on the correlation between the PM2.5 data sequence and the ozone data sequence.
In the present embodiment, the pearson coefficient between the PM2.5 data sequence and the ozone data sequence is calculated, and in order to overcome the influence when the pearson coefficient is negative, an inverse proportion value of the absolute value of the pearson coefficient is taken as the third weight. The larger the third weight value is, the smaller the correlation between the PM2.5 data sequence and the ozone data sequence is, the larger the noise influence degree of ozone is, and the larger the size of the first filtering window is determined subsequently. The process of calculating the pearson coefficients is prior art and is not within the scope of the present invention and will not be described in detail herein.
It should be noted that, the pearson coefficient may represent a correlation between PM2.5 data and ozone data, specifically expressed in: ozone and PM2.5 are homologous to pollutants, and when the emission of the pollution sources is increased, PM2.5 and ozone can be increased simultaneously, so that the change characteristics of the ozone and PM2.5 are similar, and a proportional relationship exists between the ozone and the PM 2.5; second, there may be a mutual conversion between some of the substances in PM2.5 and the chemical components in ozone, where there may be an inverse relationship between the two, but the data change characteristics are more similar. Therefore, the larger the absolute value of the pearson coefficient, the larger the correlation between the PM2.5 data and the ozone data, the smaller the ozone data is affected by noise, and the smaller the window size, so the absolute value of the pearson coefficient and the third weight are in a negative correlation.
And a fourth sub-step of combining the first weight, the second weight and the third weight to obtain an initial window size weight and determining a first filter window size of the ozone data sequence.
In this embodiment, the product of the first weight, the second weight and the third weight is calculated, normalization processing is performed on the product of the first weight, the second weight and the third weight, a normalized numerical value is obtained, and the normalized numerical value is used as an initial window size weight; calculating the product of the initial window size weight and the preset initial filter window size, upwardly rounding the product of the initial window size weight and the preset initial filter window size, and taking the upwardly rounded numerical value as the first filter window size of the ozone data sequence. The preset initial filter window size may be set to 30, and the practitioner may set the initial filter window size according to a specific practical situation, without specific limitation.
As an example, the calculation formula of the first filter window size of the ozone data sequence may be:
the method comprises the steps of carrying out a first treatment on the surface of the Wherein R is the first filter window size of the ozone data sequence, norm is a linear normalization function, < ->For the data change characteristic value corresponding to the ozone data sequence, < >>Data change characteristic value corresponding to PM2.5 data sequence, +.>For absolute function +.>First weight for initial window size, +.>For the maximum concentration value in the ozone data sequence, < >>Is the minimum concentration value in the ozone data sequence, < >>For the second weight of the initial window size exp is an exponential function based on natural constants, ++>For ozone data sequence, +.>For the PM2.5 data sequence, +.>Is the pearson coefficient between the PM2.5 data sequence and the ozone data sequence, +.>Third weight for initial window size, +.>In order to round the symbol up,and A is the preset initial filter window size for the initial window size weight.
Thus far, the present embodiment obtains a first filter window size for the ozone data sequence.
S3, determining the noise influence degree of each ozone data according to the concentration value of each ozone data in the ozone data sequence and the data change characteristic value corresponding to the preset data interval of each ozone data.
It should be noted that, in the ozone data collection process, external factors, such as wind speed and temperature, received at different times may be different, which may indicate that the noise influence degree received by each ozone data in the ozone data sequence may be different, so after the first filter window size generated by the overall noise influence of the ozone data is obtained, the first filter window size needs to be corrected according to the analysis of the local noise influence degree of the ozone data to obtain the optimal filter window size. First, it is necessary to determine the noise impact level of each ozone data in the ozone data series, each ozone data having its corresponding noise impact level.
In this embodiment, any one ozone data is used as the target ozone data, the noise influence degree of the target ozone data is determined, and a flowchart for determining the noise influence degree of the target ozone data is shown in fig. 2, and the specific implementation process includes:
first, analyzing the same concentration value number according to the concentration value of each ozone data, and determining a first noise influence factor of the target ozone data.
In this embodiment, the number of concentration values identical to the concentration value of the target ozone data in the ozone data sequence is first determined, then the number of concentration values is subjected to a reversal operation to obtain an inverse proportion value of the number of concentration values, and finally the inverse proportion value of the number of concentration values is used as the first noise influencing factor.
When the target ozone data is affected by noise, the concentration value of the target ozone data will change greatly, the noise will not have regularity, and the more the number of the ozone data sequences is the same as the concentration value of the target ozone data, the smaller the degree of the influence of the noise on the target ozone data, that is, the fewer the number of the same concentration values is, the greater the noise influence degree of the target ozone data is.
And a second step of determining a second noise influence factor of the target ozone data according to the data change characteristic values corresponding to the preset data intervals of each ozone data.
In the present embodiment, in order to quantify the local feature difference of the target ozone data and the front and rear ozone data adjacent thereto, it is necessary to determine the second noise influence factor. The larger the local feature difference, the more serious the target ozone data is affected by noise. Therefore, one of the indexes that can be used to measure the degree of influence of noise, the second noise influence factor, can be determined based on the data change characteristic values corresponding to the local areas of the adjacent ozone data.
A first sub-step of determining a preset data interval of each ozone data.
In this embodiment, for any one ozone data, a preset number of ozone data nearest to the ozone data is taken as adjacent ozone data corresponding to the ozone data, and a data section composed of the ozone data and a plurality of adjacent ozone data corresponding thereto is taken as a preset data section corresponding to the ozone data, thereby obtaining preset data sections of each ozone data. The number of adjacent ozone data may be set to 10, and the practitioner may set the number of adjacent ozone data according to the specific practical situation, which is not particularly limited herein.
It should be noted that, the preset data interval is determined to determine a local area corresponding to each ozone data, which is helpful for subsequent analysis of the data variation characteristic difference between local areas corresponding to adjacent ozone data, so as to determine the noise influence degree of the ozone data.
And a second sub-step of determining a second noise influence factor according to the data change characteristic value corresponding to the preset data interval of the target ozone data and the data change characteristic values corresponding to the preset data intervals of the previous and the next ozone data.
In this embodiment, the calculation process of the data change feature value corresponding to the target gas data sequence is determined by referring to the first step of the step S2, and the data change feature value corresponding to the preset data interval of the target ozone data and the previous and next ozone data is determined; calculating the average value of the data change characteristic values corresponding to the preset data interval of the previous ozone data and the next ozone data of the target ozone data, and taking the difference between the data change characteristic value corresponding to the preset data interval of the target ozone data and the average value of the data change characteristic values as a second noise influence factor.
And thirdly, analyzing the difference of the data change degree according to the concentration value of each ozone data, and determining a third noise influence factor of the target ozone data.
In this embodiment, a concentration value average value corresponding to the ozone data series is calculated from the concentration values of the respective ozone data, and the difference between the concentration value of the target ozone data and the concentration value average value corresponding to the ozone data series is used as the third noise influence factor. Wherein the difference between the concentration value and the concentration value mean can be determined by the absolute value of the difference between the two data.
It should be noted that, the third noise influence factor may represent the difference between the data change degree of the target ozone data and the whole change degree of the ozone data, and the larger the third noise influence factor, the larger the noise influence degree of the target ozone data in the ozone data sequence is.
Fourth, the first noise influence factor, the second noise influence factor and the third noise influence factor are fused, and the noise influence degree of the target ozone data is determined.
In this embodiment, the fusion of the three noise influence factors is achieved by calculating the product of the first noise influence factor, the second noise influence factor and the third noise influence factor, and the noise influence degree of the target ozone data is determined. Of course, the practitioner may use other ways to implement fusion of different data, which is not specifically limited herein.
As an example, the calculation formula of the noise influence degree of the i-th ozone data in the ozone data series may be:
the method comprises the steps of carrying out a first treatment on the surface of the In (1) the->For the degree of noise influence of the ith ozone data in the ozone data sequence, e is a natural constant,/->Is the same concentration value number as the concentration value of the ith ozone data in the ozone data sequence,/concentration value number>A first noise influencing factor for the ith ozone data in the ozone data sequence,/for the first noise influencing factor for the ith ozone data in the ozone data sequence>Is the data change characteristic value corresponding to the preset data interval of the ith ozone data in the ozone data sequence,/for the preset data interval of the ith ozone data in the ozone data sequence>Is the data change characteristic value corresponding to the preset data interval of the ith-1 th ozone data in the ozone data sequence,/and%>Is the data change characteristic value corresponding to the preset data interval of the (i+1) th ozone data in the ozone data sequence,/and%>A second noise influencing factor for the ith ozone data in the ozone data sequence,/for the ozone data sequence>Concentration value of the ith ozone data in the ozone data series,/for the ozone data>For the mean value of the concentration values corresponding to the ozone data sequence, < + >>Third noise influencing factor for the ith ozone data in the ozone data sequence, +.>For absolute value functions.
It should be noted that, the calculation formula of the noise influence degree of the ith ozone data cannot be used for determining the noise influence degree of the 1 st ozone data in the ozone data sequence, so that the 1 st ozone data in the omitted ozone data sequence can be determined without adaptive filter window, and the 1 st ozone data in the ozone data sequence can be filtered by using a filter window with a traditional fixed window size.
Up to this point, the present embodiment obtains the noise influence degree of each ozone data in the ozone data sequence.
And S4, correcting the size of the first filter window according to the noise influence degree of each ozone data, and determining the optimal filter window size of each ozone data.
It should be noted that, the filter window in the filter algorithm is not only affected by the whole noise, but also affected by the local noise, that is, different ozone data are affected by noise, and the corresponding filter windows should also be different, so that the noise affected degree of each ozone data can be used to correct the first filter window size, so as to obtain the optimal filter window size of each ozone data, and each ozone data has its corresponding optimal filter window size.
In this embodiment, taking any one selected ozone data of all ozone data as an example, determining an optimal filter window size of the ozone data, the specific implementation process includes:
first, determining a fourth weight according to the noise influence degree of each ozone data. In this embodiment, an average value of noise influence degrees of all ozone data is first determined, which can be used to measure the noise influence degree of the whole ozone data; calculating the ratio of the noise influence degree of the ozone data to the average value of the noise influence degrees of all the ozone data, wherein the ratio of the noise influence degree of the single ozone data to the overall noise influence degree can be quantized; and taking the ratio as a fourth weight when determining the optimal filter window size corresponding to the ozone data.
It should be noted that, the fourth weight may represent the specific gravity of the noise influence degree of the selected ozone data relative to other ozone data, and the larger the fourth weight, the larger the noise influence degree of the selected ozone data is, the larger the ratio of the noise influence degree of the selected ozone data is, which may need to be filtered by a larger filtering window, so as to ensure the final filtering effect; otherwise, a smaller filter window is required to avoid excessive filtering and loss of data change detail.
And a second step of determining a fifth weight according to the data change characteristic value corresponding to the preset data interval of the ozone data and the data change characteristic value corresponding to the ozone data sequence.
In this embodiment, a ratio of a data change characteristic value corresponding to a preset data interval of ozone data to a data change characteristic value corresponding to an ozone data sequence is used as the fifth weight.
It should be noted that, the fifth weight may reflect the difference between the local variation characteristic of the selected ozone data and the overall data variation characteristic, and when the local variation characteristic of the single ozone data is greater than the variation characteristic of the overall data, a filtering window with a larger size should be selected to ensure that when the selected ozone data is filtered, the filtering window with a larger size is considered, so that the accuracy of filtering the ozone data can be ensured.
And thirdly, determining a final window size weight according to the fourth weight and the fifth weight.
In this embodiment, the product of the fourth weight and the fifth weight is calculated, normalized, and the normalized value of the product of the fourth weight and the fifth weight is obtained and determined as the final window size weight.
It should be noted that the value range of the final window size weight is between 0 and 1. The fourth weight and the fifth weight are both positively correlated with the final window size weight, and the larger the fourth weight and the fifth weight, the larger the final window size weight.
And step four, determining the optimal filter window size of the ozone data according to the final window size weight, the preset adjusting parameters and the first filter window size.
In this embodiment, the product of the final window size weight, the preset adjustment parameter and the first filter window size is rounded up, and the rounded up value is used as the optimal filter window size of the ozone data. The preset adjustment parameter may represent an adjustment parameter of the optimal filter window size of each ozone data during filtering, may be set to 2, and an implementer may set the preset adjustment parameter according to a specific practical situation, which is not specifically limited herein.
As an example, the calculation formula of the optimal filter window size of the ith ozone data in the ozone data sequence may be:
the method comprises the steps of carrying out a first treatment on the surface of the In (1) the->For the optimal filter window size of the ith ozone data in the ozone data sequence, norm is a linear normalization function, +.>Is the data change characteristic value corresponding to the preset data interval of the ith ozone data in the ozone data sequence,/for the preset data interval of the ith ozone data in the ozone data sequence>For the data change characteristic value corresponding to the ozone data sequence, < >>For determining the fifth weight value of the optimal filter window size corresponding to the ith ozone data,/->For the noise influence level of the ith ozone data in the ozone data sequence, +.>For the mean value of the noise influence degree of all ozone data, +.>For all ozone data number, +.>In order to determine a fourth weight value when the optimal filter window size corresponding to the ith ozone data is determined, X is a preset adjusting parameter, and R isFirst filter window size,/>To round the symbol up.
Thus far, the present embodiment obtains the optimal filter window size for each ozone data.
And S5, carrying out filtering treatment on the ozone data sequence according to the optimal filtering window size of each ozone data to obtain a filtered ozone data sequence.
In this embodiment, for any one ozone data in the ozone data sequence, a filter window is determined by using an optimal filter window size of the ozone data, and a non-local bilateral filter process is performed on the ozone data by using the filter window, so that a filter result of the ozone data can be obtained. Each ozone data in the ozone data sequence is subjected to the filtering processing, so that a filtering result of each ozone data can be obtained, and a sequence formed by the filtering results of all the ozone data is used as the ozone data sequence after the filtering processing. The implementation process of the non-local bilateral filtering process is the prior art and is not within the scope of the present disclosure, and will not be described in detail herein.
When the non-local bilateral filtering algorithm is used for denoising the ozone data, the size of the filtering window can influence the accuracy of removing the ozone noise finally, and the embodiment utilizes the filtering window with the determined optimal filtering window size of each ozone data to remove the ozone noise, so that the method not only can adapt to different noise influence degrees of each ozone data, but also can improve the filtering effect and obtain the filtered ozone data sequence with higher accuracy and stronger authenticity.
S6, determining the prediction results of the PM2.5 and the ozone by using a prediction model according to the PM2.5 data sequence and the filtered ozone data sequence.
In this embodiment, the PM2.5 data sequence and the filtered ozone data sequence are used as input data, and are input into an ARIMA prediction model which is built and trained in advance, so that a prediction result of PM2.5 and ozone can be obtained. The accuracy of the ozone data sequence after the filtering processing determined by the optimal filtering window size is higher, namely the accuracy of input data is improved, the reliability of the output result of the ARIMA prediction model can be effectively improved by the accurate input data, and the accuracy of the air pollution prevention and control prediction analysis based on PM2.5 and ozone data is further improved. The construction and training process of the ARIMA prediction model is the prior art and is not within the scope of the present invention, and will not be described in detail herein.
It should be noted that, the obtained prediction result may be PM2.5 data and ozone data at a certain time point in the future, or PM2.5 data and ozone data at a certain time point in the future, and if the PM2.5 data and ozone data at a certain time point in the future, the concentration change trend of the predicted gas pollutant may be determined according to the PM2.5 data and ozone data in the time point. The specific data form of the prediction result can be set by an implementer according to the specific actual situation, and the parameters of the ARIMA prediction model are not particularly limited.
After the predicted results of PM2.5 and ozone are obtained, corresponding air pollution prevention and control measures can be adopted later according to the predicted results, for example, when the concentrations of ozone and PM2.5 in a future period of time are predicted to be higher, measures such as energy conservation and emission reduction are formulated, for example, the running limit of an automobile, the working limit of a factory and the like are formulated, so that pollutant emission amount is reduced or pollutant emission time is staggered.
The embodiment of the invention also provides a prediction system for PM2.5 and ozone pollution cooperative prevention and control, which comprises a processor and a memory, wherein the processor is used for processing instructions stored in the memory so as to realize the prediction method for PM2.5 and ozone pollution cooperative prevention and control.
According to the invention, the window size of the ozone data is adaptively determined when the non-local bilateral filtering is used, the accuracy of filtering the collected ozone data is improved, and the accuracy of air pollution prevention and control prediction analysis based on PM2.5 and ozone data is further improved.
An embodiment of an ozone data processing method for pollution cooperative prevention and control prediction comprises the following steps:
with the popularization of industry and modern transportation means, the generated waste brings certain pollution to the environment, PM2.5 and ozone in the waste are main atmospheric pollutants in urban areas, wherein the ozone is secondary pollutant, PM2.5 has the properties of primary pollutant and secondary pollutant, and a complex interaction process exists between the two. PM2.5 and ozone in the air are monitored and predicted, and the method plays an important role in environmental management, prevention and control. When PM2.5 and ozone in the air are monitored and predicted, noise data exists in ozone data monitored by a monitor because the ozone is more easily influenced by factors in the surrounding environment, such as illumination, temperature, air flow and the like, and the authenticity of the ozone data is difficult to guarantee. PM2.5 in air is relatively stable compared to ozone. Therefore, filtering processing is required for ozone data.
The conventional common non-local bilateral filtering algorithm carries out filtering treatment on ozone data, the filtering treatment is realized by carrying out local analysis on a window with a fixed size during filtering, the selection of the size of the window has a certain influence on the final filtering effect, the fixed size filtering window set in the conventional algorithm cannot adapt to different noise influence degrees of different data, the final filtering effect of the ozone data is promoted to be poor, the accuracy of the ozone data processing result is low, and the subsequent PM2.5 and filtered ozone data-based air pollution prevention and control prediction analysis is not facilitated.
In order to solve the technical problem that the accuracy of the ozone data processing result is low because the filtering window of the existing non-local bilateral filtering algorithm cannot adapt to different noises in the ozone data, the embodiment provides an ozone data processing method for pollution collaborative prevention and control prediction, which comprises the following steps:
s1, acquiring a PM2.5 data sequence and an ozone data sequence;
s2, correcting the preset initial filter window size according to the PM2.5 data sequence and the ozone data sequence, and determining the first filter window size of the ozone data sequence;
s3, determining the noise influence degree of each ozone data according to the concentration value of each ozone data in the ozone data sequence and the data change characteristic value corresponding to the preset data interval of each ozone data;
S4, correcting the size of the first filter window according to the noise influence degree of each ozone data, and determining the optimal filter window size of each ozone data;
and S5, carrying out filtering treatment on the ozone data sequence according to the optimal filtering window size of each ozone data to obtain a filtered ozone data sequence.
The ozone data processing method for pollution cooperative prevention and control prediction provided by the embodiment has the following technical effects:
because collected ozone data is easy to be influenced by surrounding environment to cause noise in the ozone data, in order to improve the accuracy of filtering and denoising of the ozone data and further improve the accuracy of an air pollution prevention and control prediction analysis result, the embodiment combines the data characteristics of PM2.5 data and ozone data, quantifies the noise influence degree of each ozone data, and adaptively determines the window size of the ozone data when non-local bilateral filtering is used, namely determines the optimal filtering window size of each ozone data. Firstly, when ozone data is changed, PM2.5 data also changes to a certain extent, so that when the first filter window size of the ozone data sequence is analyzed, the numerical accuracy of the first filter window size can be effectively improved by combining the data characteristics of the PM2.5 data; secondly, when the size of the first filtering window is determined, not only the data change characteristic values corresponding to the PM2.5 data sequence and the ozone data sequence are considered, but also the maximum concentration value and the minimum concentration value in the ozone data sequence and the correlation between the PM2.5 data sequence and the ozone data sequence are considered, and the influence degree of noise on the whole ozone data is quantized from multiple angles, so that the reliability degree of the size of the first filtering window is improved; then, the local noise influence in the ozone data sequence is further quantized, namely, the noise influence degree of each ozone data is determined, the noise influence degree of each ozone data is utilized to correct the first filter window size, each ozone data can be caused to have the corresponding adaptive optimal filter window size, compared with the size of the filter window which is directly and manually set, the optimal filter window size of each ozone data determined by the embodiment can adapt to ozone data with different noise influence degrees, the filter effect of the ozone data is improved, and ozone data with higher accuracy is obtained; based on ozone data and PM2.5 data with higher accuracy, the method is beneficial to improving the accuracy of the follow-up pollution collaborative prevention and control prediction.
The steps S1-S5 are already described in detail in the embodiment of the method and the system for predicting the cooperative prevention and control of PM2.5 and ozone pollution, and are not described in detail.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
Claims (10)
1. The prediction method for PM2.5 and ozone pollution cooperative prevention and control is characterized by comprising the following steps of:
acquiring a PM2.5 data sequence and an ozone data sequence; according to the PM2.5 data sequence and the ozone data sequence, determining data change characteristic values corresponding to the PM2.5 data sequence and the ozone data sequence;
correcting a preset initial filter window size according to the data change characteristic value corresponding to the PM2.5 data sequence and the ozone data sequence, the maximum concentration value and the minimum concentration value in the ozone data sequence and the correlation between the PM2.5 data sequence and the ozone data sequence, and determining a first filter window size of the ozone data sequence;
Determining the noise influence degree of each ozone data according to the concentration value of each ozone data in the ozone data sequence and the data change characteristic value corresponding to the preset data interval of each ozone data;
correcting the first filter window size according to preset adjustment parameters, noise influence degree of each ozone data, data change characteristic values corresponding to preset data intervals and data change characteristic values corresponding to ozone data sequences, and determining the optimal filter window size of each ozone data;
according to the optimal filter window size of each ozone data, carrying out filter treatment on the ozone data sequence to obtain a filtered ozone data sequence; and determining the prediction results of the PM2.5 and the ozone by using a prediction model according to the PM2.5 data sequence and the filtered ozone data sequence.
2. The method for predicting cooperative prevention and control of PM2.5 and ozone pollution according to claim 1, wherein the determining the characteristic value of the data change corresponding to the PM2.5 data sequence and the ozone data sequence according to the PM2.5 data sequence and the ozone data sequence comprises:
taking the PM2.5 data sequence or the ozone data sequence as a target gas data sequence, determining standard deviation and mean value of the target gas data sequence, and taking the ratio of the standard deviation to the mean value as a first variation characteristic factor;
Calculating the corresponding slope of the adjacent gas data according to each gas data in the target gas data sequence to obtain each slope; determining differences between adjacent slopes according to the slopes to obtain the differences of the slopes, and taking the average value of the differences of all the slopes as a second change characteristic factor;
and combining the first change characteristic factor and the second change characteristic factor to obtain a data change characteristic value corresponding to the target gas data sequence.
3. The method for predicting cooperative prevention and control of PM2.5 and ozone pollution according to claim 1, wherein the step of correcting the preset initial filter window size according to the correlation between the data change characteristic value corresponding to the PM2.5 data sequence and the ozone data sequence, the maximum concentration value and the minimum concentration value in the ozone data sequence, and the PM2.5 data sequence and the ozone data sequence to determine the first filter window size of the ozone data sequence comprises:
determining the difference between the data change characteristic value corresponding to the PM2.5 data sequence and the data change characteristic value corresponding to the ozone data sequence, and taking the difference between the two data change characteristic values as a first weight;
obtaining a maximum concentration value and a minimum concentration value in an ozone data sequence, calculating a difference value between the maximum concentration value and the minimum concentration value, and taking the difference value between the maximum concentration value and the minimum concentration value as a second weight;
Determining a pearson coefficient according to the PM2.5 data sequence and the ozone data sequence, and taking an inverse proportion value of the absolute value of the pearson coefficient as a third weight;
and combining the first weight, the second weight and the third weight to obtain an initial window size weight, upwardly rounding the product of the initial window size weight and a preset initial filter window size, and taking the upwardly rounded value as the first filter window size of the ozone data sequence.
4. A method for predicting a synergistic PM2.5 and ozone pollution control as claimed in claim 3, wherein said combining the first weight, the second weight and the third weight to obtain the initial window size weight comprises:
calculating the product of the first weight, the second weight and the third weight, normalizing the product of the first weight, the second weight and the third weight to obtain a normalized value, and taking the normalized value as an initial window size weight.
5. The method for predicting cooperative prevention and control of PM2.5 and ozone pollution according to claim 1, wherein the determining the noise influence degree of each ozone data according to the concentration value of each ozone data in the ozone data sequence and the data change characteristic value corresponding to the preset data interval of each ozone data comprises:
Determining the number of concentration values which are the same as the concentration value of the target ozone data in an ozone data sequence by taking any one of the ozone data as the target ozone data, performing inverse operation on the number of concentration values to obtain an inverse proportion value of the number of concentration values, and taking the inverse proportion value of the number of concentration values as a first noise influence factor;
determining preset data intervals of each ozone data, and determining a second noise influence factor according to the data change characteristic value corresponding to the preset data interval of the target ozone data and the data change characteristic values corresponding to the preset data intervals of the previous and next ozone data;
calculating a concentration value average value corresponding to the ozone data sequence, and taking the difference between the concentration value of the target ozone data and the concentration value average value corresponding to the ozone data sequence as a third noise influence factor;
and fusing the first noise influence factor, the second noise influence factor and the third noise influence factor, and determining the noise influence degree of the target ozone data.
6. The method for predicting PM2.5 and ozone pollution co-control according to claim 5, wherein determining the second noise impact factor according to the data change characteristic value corresponding to the preset data interval of the target ozone data and the data change characteristic values corresponding to the preset data intervals of the preceding and following ozone data comprises:
Calculating the average value of the data change characteristic values corresponding to the preset data interval of the previous ozone data and the next ozone data of the target ozone data, and taking the difference between the data change characteristic value corresponding to the preset data interval of the target ozone data and the average value of the data change characteristic values as a second noise influence factor.
7. The method for predicting cooperative prevention and control of PM2.5 and ozone as set forth in claim 6, wherein the predetermined data interval of ozone data is a data interval of a predetermined number of ozone data nearest to the ozone data and itself.
8. The method for predicting cooperative prevention and control of PM2.5 and ozone pollution according to claim 1, wherein the determining the optimal filter window size of each ozone data by correcting the first filter window size according to the preset adjustment parameter, the noise influence degree of each ozone data, the data change characteristic value corresponding to the preset data interval, and the data change characteristic value corresponding to the ozone data sequence comprises:
determining an average value of noise influence degrees of all ozone data; for any one ozone data, determining the ratio between the noise influence degree of the ozone data and the average value of the noise influence degrees of all the ozone data as a fourth weight;
Taking the ratio of the data change characteristic value corresponding to the preset data interval of the ozone data to the data change characteristic value corresponding to the ozone data sequence as a fifth weight;
calculating the product of the fourth weight and the fifth weight, normalizing the product of the fourth weight and the fifth weight to obtain a normalized value of the product of the fourth weight and the fifth weight, and determining the normalized value as a final window size weight;
and (3) carrying out upward rounding on the product of the final window size weight, the preset adjusting parameter and the first filter window size, and taking the numerical value obtained after upward rounding as the optimal filter window size of the ozone data.
9. The method for predicting PM2.5 and ozone pollution co-control according to claim 1, wherein the determining the predicted result of PM2.5 and ozone according to the PM2.5 data sequence and the filtered ozone data sequence using a prediction model comprises:
and (3) taking the PM2.5 data sequence and the filtered ozone data sequence as input data, and inputting the input data into a pre-constructed and trained ARIMA prediction model to obtain a PM2.5 and ozone prediction result.
10. A prediction system for co-control of PM2.5 and ozone pollution, comprising a processor and a memory, the processor being configured to process instructions stored in the memory to implement a prediction method for co-control of PM2.5 and ozone pollution as claimed in any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410057175.XA CN117574061B (en) | 2024-01-16 | 2024-01-16 | PM2.5 and ozone pollution cooperative prevention and control prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410057175.XA CN117574061B (en) | 2024-01-16 | 2024-01-16 | PM2.5 and ozone pollution cooperative prevention and control prediction method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117574061A true CN117574061A (en) | 2024-02-20 |
| CN117574061B CN117574061B (en) | 2024-04-05 |
Family
ID=89864795
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410057175.XA Active CN117574061B (en) | 2024-01-16 | 2024-01-16 | PM2.5 and ozone pollution cooperative prevention and control prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117574061B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109034252A (en) * | 2018-08-01 | 2018-12-18 | 中国科学院大气物理研究所 | The automatic identification method of air quality website monitoring data exception |
| CN110361505A (en) * | 2019-07-25 | 2019-10-22 | 中南大学 | A kind of outer atmosphere pollution environment Train occupant health early warning system of vehicle and its method |
| US20230012177A1 (en) * | 2021-07-07 | 2023-01-12 | The Bank Of New York Mellon | System and methods for generating optimal data predictions in real-time for time series data signals |
| CN116679356A (en) * | 2023-05-22 | 2023-09-01 | 湖南大学 | Method for predicting PM2.5 and ozone mixed pollution under high space-time resolution |
-
2024
- 2024-01-16 CN CN202410057175.XA patent/CN117574061B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109034252A (en) * | 2018-08-01 | 2018-12-18 | 中国科学院大气物理研究所 | The automatic identification method of air quality website monitoring data exception |
| CN110361505A (en) * | 2019-07-25 | 2019-10-22 | 中南大学 | A kind of outer atmosphere pollution environment Train occupant health early warning system of vehicle and its method |
| US20230012177A1 (en) * | 2021-07-07 | 2023-01-12 | The Bank Of New York Mellon | System and methods for generating optimal data predictions in real-time for time series data signals |
| CN116679356A (en) * | 2023-05-22 | 2023-09-01 | 湖南大学 | Method for predicting PM2.5 and ozone mixed pollution under high space-time resolution |
Non-Patent Citations (3)
| Title |
|---|
| CHUNSHENG FANG ET AL.: "Analysis of the meteorological impact on PM2.5 pollution in Changchun based on KZ filter and WRF-CMAQ", LONG-TERM TREND OF OZONE IN SOUTHERN CHINA REVEALS FUTURE MITIGATION STRATEGY FOR AIR POLLUTION, 26 December 2021 (2021-12-26), pages 1 - 15 * |
| XIAO-BING LI ET AL.: "Long-term trend of ozone in southern China reveals future mitigation strategy for air pollution", ATMOSPHERIC ENVIRONMENT, 24 November 2021 (2021-11-24), pages 1 - 9 * |
| 郑海明 等: "差分吸收光谱测量臭氧浓度信号去噪的实验研究", 计量学报, no. 06, 28 June 2020 (2020-06-28), pages 130 - 135 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117574061B (en) | 2024-04-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111967688B (en) | Power load prediction method based on Kalman filter and convolutional neural network | |
| CN115933787B (en) | Indoor multi-terminal intelligent control system based on indoor environment monitoring | |
| CN112465239B (en) | Desulfurization system operation optimization method based on improved PSO-FCM algorithm | |
| CN115962551B (en) | Intelligent air conditioner control system and method for building automatic control | |
| CN116441031B (en) | Intelligent crushing system for garbage incineration slag | |
| CN114329347B (en) | Method and device for predicting metering error of electric energy meter and storage medium | |
| CN117574061B (en) | PM2.5 and ozone pollution cooperative prevention and control prediction method and system | |
| CN117368141B (en) | Perchlorate wastewater concentration intelligent detection method based on artificial intelligence | |
| CN117407700A (en) | Method for monitoring working environment in live working process | |
| CN116596139A (en) | Short-term load prediction method and system based on Elman neural network | |
| CN113887119A (en) | River water quality prediction method based on SARIMA-LSTM | |
| CN115099291B (en) | Building energy-saving monitoring method | |
| CN115829157A (en) | Chemical water quality index prediction method based on variational modal decomposition and auto former model | |
| CN115459868A (en) | Millimeter wave communication performance evaluation method and system in complex environment | |
| CN117786325B (en) | Ambient temperature wisdom early warning system of heavy-calibre thing networking water gauge | |
| CN113866204A (en) | Bayesian regularization-based soil heavy metal quantitative analysis method | |
| CN113222929A (en) | Smoke concentration detection method and device based on total variation | |
| CN117670000B (en) | Pump station water supply quantity prediction method based on combined prediction model | |
| CN117435873B (en) | Data management method based on intelligent spraying dust fall | |
| CN117434153B (en) | Road nondestructive testing method and system based on ultrasonic technology | |
| CN114626008B (en) | Railway subgrade settlement prediction method and device based on power-related random process | |
| CN113204831B (en) | Design method of dynamic baseline of ship system equipment | |
| CN112906101B (en) | Bridge residual deformation abnormity assessment early warning method based on monitoring data | |
| CN116823067B (en) | Method and device for determining water quality cleaning state of pipe network and electronic equipment | |
| CN113878613B (en) | Industrial robot harmonic reducer early fault detection method based on WLCTD and OMA-VMD |
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 |