Ozone concentration forecasting method based on peak perception and influence factor analysis

Abstract

The invention discloses an ozone concentration forecasting method based on peak value perception and influence factor analysis, which utilizes continuous online observation data and meteorological observation data of multi-site pollutants, processes abnormal and missing data, and combines with boundary layer height, radiation and sea level air pressure provided by an ECMWF meteorological forecasting product to forecast ozone concentration. Building a high-quality Yunnan region large database for supporting deep learning according to hour-scale meteorological numerical simulation data of vertical wind speed, and building an ozone forecasting model based on influence factor analysis and peak value perception; capturing a global dependency relationship among historical meteorology, pollutant data and an ozone long-time sequence and a nonlinear relationship between ozone and an influence factor, so that the model can better understand context information and a related mechanism influencing ozone generation; the defect of long-time sequence dependence of a traditional deep learning method is overcome, the influence of influence factors on ozone concentration prediction is ignored, and the accuracy of ozone prediction is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental prediction, and in particular to an ozone concentration prediction method based on peak perception and influencing factor analysis. Background Art

[0002] Ozone is a strong oxidant. Its presence in the atmosphere can not only cause harm to human health, but also have a negative impact on crops and ecosystems. Traditional ozone concentration forecasting methods often rely on statistical models and empirical formulas, which have limitations in dealing with the complex nonlinear relationship and long time series data between ozone and its influencing factors. Therefore, in order to solve the above problems, this paper proposes an ozone concentration forecasting method based on peak perception and influencing factor analysis. Summary of the invention

[0003] In order to solve the above technical problems, the present invention designs an ozone concentration prediction method based on peak perception and influencing factor analysis.

[0004] In order to achieve the above technical effects, the present invention is implemented by the following technical solutions: a method for predicting ozone concentration based on peak perception and influencing factor analysis, characterized in that it includes the following steps:

[0005] S1. Obtain the monitoring data of pollutants and their meteorological hourly scale in Yunnan for many years through network data collection, and combine ECMWF meteorological forecast products to provide hourly scale data of boundary layer height, radiation, sea level pressure, and vertical wind speed. Then, perform outlier processing, null value filling, data normalization, and data set division on the pollutant and meteorological monitoring data to build a large database of pollutants and meteorological data in Yunnan that supports deep learning.

[0006] S2. Use the Transformer model based on influencing factor analysis and peak perception optimization to train the historical observation data in the Yunnan pollutant and meteorological data database and the additional meteorological parameter data set provided by ECMWF, establish the Yunnan ozone forecast model based on historical meteorology, pollutants and future ozone, and optimize the Yunnan ozone forecast model through peak perception and influencing factor analysis based on deep learning, and output n-master control factors;

[0007] S3. The evaluation results of the optimized Yunnan ozone forecast model in the validation set are used as indicators to evaluate the quality of the deep learning model, and the ozone concentration is evaluated; the evaluation indicators include mean square error (MSE), mean absolute error (MAE), and determination coefficient (R). 2 , select evaluation indicators

[0008] Metric=(MSE+MAE) / 2-R 2As an evaluation indicator.

[0009] The model with the lowest metric on the validation set is selected as the optimal model and used for fast and accurate forecasting; where MSE is mean square error, MAE is absolute error, and R 2 is the determination coefficient.

[0010] Furthermore, in S1, a large database of pollutants and meteorological data in Yunnan is established as follows:

[0011] Through online data download, the six conventional monitoring pollutants (O3, CO, SO2, NO2, PM 2.5 、PM 10 ) Continuous online observation data and meteorological observation data are combined with ECMWF meteorological forecast products to provide hourly-scale meteorological numerical simulation data on boundary layer height, radiation, sea level pressure, and vertical wind speed to collect and organize, and build a long-term historical data set of ozone, meteorology, and other pollutants; and perform data outlier processing, null value filling, data normalization, and data set division on the data set to meet the data format required for deep learning, and establish a large database of pollutants and meteorological data in Yunnan to support deep learning;

[0012] During the project implementation period, meteorological and pollutant observation data will be continuously stored and sorted to achieve dynamic updating of the large database.

[0013] Furthermore, the data outlier processing, null value filling, data normalization and data set division are specifically as follows:

[0014] Data outlier processing: abnormal reasons such as equipment damage may cause the monitoring data to be too large or the equipment to be insensitive, resulting in the monitoring data being continuously 0 for a period of time. This part of the data content in the conventional monitoring pollutant data and meteorological data is eliminated;

[0015] Filling in empty values: Due to observation equipment or observation conditions, the phenomenon of observation data discontinuity is obvious, so it is very important to supplement missing values. The data filling method adopted in the present invention is a filling method based on statistical value interpolation at different time scales:

[0016] V ij =|V monthij +V weekdayij +V hourij -2×V yearij |

[0017] Among them, V ij is the filling value of pollutant i at missing time j, Vmonth ij is the mean concentration of pollutant i in the month of missing time j, Vweekdayij is the mean concentration of pollutant i in the week of missing time j, Vyeay ij is the mean pollutant concentration of pollutant i in the year of missing time j;

[0018] Data normalization: Data normalization is conducive to improving the stability of training and the convergence speed of model training. The present invention adopts the normal distribution normalization method, which is to subtract the mean of the characteristic attributes of the data and divide it by the variance; convert it into a standard normal distribution with a mean of 0 and a variance of 1:

[0019] x=(xv) / σ

[0020] In the formula, μ is the data mean; σ is the data variance; record the mean and variance on the training set, standardize the training set, and use the variance and mean on the training set to standardize the test set when using the model in subsequent tests;

[0021] Dataset division: The data used are from more than 100 ground observation stations in the Yunnan monitoring network starting in 2015 to monitor O3, NO2, CO, SO2, PM 2.5 , PM 10 Hourly continuous observation data and ECMWF weather forecast products provide hourly-scale meteorological numerical simulation data of boundary layer height, radiation, sea level pressure, and vertical wind speed. Data from the past year are selected as the test set to evaluate the model capability, and other data are divided into training set and validation set according to the data volume of 4:1. The training set is used to train model parameters, and the validation set is used to select the model with the best number of training rounds.

[0022] Furthermore, in S2, based on big data deep learning, the ozone forecast model in Yunnan is optimized through peak perception and influencing factor analysis as follows:

[0023] S2.1. The Transformer model based on influencing factor analysis is used to train the historical data set. The meteorological and atmospheric pollutant monitoring data of more than ten years and the numerical simulation data provided by the ECMWF meteorological forecast products are used as the input of the neural network model, and the predicted ozone concentration is used as the output. The ozone forecast model for the Yunnan region at the hourly scale and the daily scale is established respectively; and the perception score is output, and the largest n perception scores are selected, and the corresponding influencing factors are taken as n-main control factors;

[0024] S2.2. Simultaneously apply the peak sensing method to find the corresponding peak values ​​in the daily forecast and hourly forecast results using the peak finding algorithm, including:

[0025] PeakIndex=PeakDetection(Target)

[0026] Where Target is the true value of the predicted ozone concentration, Target∈R B×L‘ , PeakIndex is the index of the peak, PeakIndex∈{1,2,3,...,L'}; PeakDetection is the peak search algorithm, here is the Automatic multiscale-based peak detection (AMPD) algorithm;

[0027] Calculate the peak perceptual loss function and perform gradient backpropagation

[0028] Loss peak =MSELoss(Targe PeakIndex ,FinalOut PeakIndex )

[0029] Target PeakIndex , indicating that by Target i The sequence of numbers, where i∈PeakIndex; similarly FinalOut PeakIndex , indicating that FinalOut i The sequence of numbers, where i∈PeakIndex;

[0030] Loss MSE =MSELoss(Target,FinalOut)

[0031] Loss total =Loss MSE +αLoss peak

[0032] In the formula, α is 1;

[0033] Then, the penalty level of the model is increased at the corresponding peak value, so that the model can better perceive the peak time and the corresponding concentration, improve the ozone peak forecast effect, and realize the rapid and accurate forecast of future severe ozone pollution process in Yunnan Province; thus, the optimization of the ozone forecast model in Yunnan Province can be completed.

[0034] Furthermore, the meteorological data set includes temperature, wind speed, wind direction, total cloud cover, relative humidity, atmospheric pressure, boundary layer height, radiation, sea level pressure and vertical wind speed.

[0035] The beneficial effects of the present invention are:

[0036] The present invention utilizes continuous online observation data of pollutants and meteorological observation data from multiple sites, processes abnormal and missing data, and combines ECMWF meteorological forecast products to provide hourly-scale meteorological numerical simulation data of boundary layer height, radiation, sea level pressure, and vertical wind speed, establishes a high-quality Yunnan region big database for supporting deep learning, and establishes an ozone forecast model based on influencing factor analysis and peak perception, captures the global dependency relationship between historical meteorological and pollutant data and ozone long-term series, and the nonlinear relationship between ozone and influencing factors, so that the model can better understand contextual information and related mechanisms affecting ozone generation, solves the defects of traditional deep learning methods in long-term series dependency and ignores the impact of influencing factors on predicted ozone concentration, and improves the accuracy of ozone forecast; by introducing the peak perception method, the accuracy of the model's ozone peak occurrence time and concentration forecast is improved, the forecast effect of the ozone heavy pollution process is optimized, and a fast and accurate forecast model suitable for the ozone heavy pollution process in Yunnan is established, providing more detailed scientific support for Yunnan Province to formulate dynamic and refined ozone peak reduction prevention and control measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0038] Figure 1 It is a flow chart of the ozone concentration forecasting method based on peak perception and influencing factor analysis described in the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] Example 1

[0041] A method for predicting ozone concentration based on peak perception and influencing factor analysis comprises the following steps:

[0042] Step 1: Collect and organize pollutant and meteorological monitoring data in Yunnan and establish a large ozone pollution database, as follows:

[0043] The accumulated data on O3, NO2, CO, SO2, and PM in Yunnan since 2015 are obtained by downloading online data.2.5 , PM 10 Hourly continuous observation data. Currently, hourly observation data from 86 meteorological observation stations can be obtained. The observed meteorological elements include temperature, wind speed, wind direction, total cloud cover, relative humidity and atmospheric pressure.

[0044] In addition, the present invention selects ECMWF numerical forecast products with a resolution of 0.125°×0.125°, and the meteorological data set includes temperature, wind speed, wind direction, total cloud cover, relative humidity and atmospheric pressure. On the basis of collecting and collating historical ground observation data and numerical forecast product output boundary layer height, radiation, sea level pressure, vertical wind speed meteorological numerical simulation data, through data outlier processing, null value filling, data normalization and data set division, so that it meets the data format required for deep learning, a large database of pollutants and meteorological data in Yunnan area for supporting deep learning is established.

[0045] Data outlier processing: Abnormal equipment damage and other reasons may cause the monitoring data to be too large, or equipment insensitivity may cause the monitoring data to be continuously 0 for a period of time. This part of the data content in the conventional monitoring pollutant data and meteorological data is eliminated.

[0046] Data completion: Due to reasons such as observation equipment or observation conditions, the phenomenon of discontinuity of observation data is obvious, so it is very important to supplement missing values. The data completion method adopted in the present invention is a completion method based on interpolation of statistical values ​​at different time scales.

[0047] V ij =|V monthij +V weekdayij +V hourij -2×V yearij |

[0048] Among them, V ij is the filling value of pollutant i at missing time j, Vmonth ij is the mean concentration of pollutant i in the month of missing time j, Vweekday ij is the mean concentration of pollutant i in the week of missing time j, Vyeay ij is the mean pollutant concentration of pollutant i in the year of missing time j.

[0049] Data normalization: Data normalization is beneficial to improving the stability of training and the convergence speed of model training. The normal distribution normalization method adopted in this invention is to subtract the mean of the characteristic attributes of the data and divide it by the variance. It is converted into a standard normal distribution with a mean of 0 and a variance of 1.

[0050] x=(x-μ) / σ

[0051] Where μ is the data mean, σ is the data variance, record the mean and variance on the training set, and after standardizing the training set, when using the model in subsequent tests, use the variance and mean on the training set to standardize the test data.

[0052] Dataset division: The data used in this paper are from 2015, when more than 100 ground observation stations in the Yunnan monitoring network began to conduct O3, NO2, CO, SO2, PM 2.5 , PM 10 Continuous observation data hour by hour.

[0053] Data from the past year was selected as the test set to evaluate the model capability, and the remaining data was divided into a training set and a validation set at a ratio of 4:1. The training set was used to train the model parameters, and the validation set was used to select the model with the best number of training rounds.

[0054] Step 2: Use the Transformer model based on influencing factor analysis and peak-aware optimization to train the historical observation data set in the Yunnan pollutant and meteorological data database, establish an ozone forecast model for Yunnan based on historical meteorology, pollutants and future ozone, and optimize the ozone forecast model for Yunnan based on deep learning, as follows:

[0055] The Transformer model based on influencing factor analysis was used to train the historical observation data set. More than a decade of meteorological and atmospheric pollutant monitoring data were used as the input of the neural network model, and the predicted ozone concentration was used as the output. The ozone forecast model for Yunnan region was established on the hourly and daily scales respectively, and the influence degree score of the influencing factor on the output predicted ozone concentration was output. The largest n influence degree scores were selected, and their corresponding influence factors were taken as the n-master controlling factors.

[0056] At the same time, the peak perception method is applied to the ozone concentration at the daily and hourly scales respectively, and the penalty level of the model is increased at the corresponding peak, so that the model can better perceive the peak time and the corresponding concentration, improve the ozone peak forecast effect, and realize the rapid and accurate forecast of future severe ozone pollution processes in Yunnan Province; thus, the optimization of the ozone forecast model in Yunnan Province can be completed.

[0057] A. Model selection principles

[0058] The self-attention mechanism is used to train historical observation data sets to accurately extract contextual features between time series, establish corresponding temporal context dependencies and interact through channel information to accurately establish the temporal dependency between ozone and meteorological and other pollutants.

[0059] B. Model to be established

[0060] Establish a scientific and effective neural network model, set the number of encoding layers and decoding layers to 1, 2, 3, 4, etc., the number of multi-head attention heads to 8, the training batch to 256, select the L2 regularization algorithm, the input data length to 120, 96, 72, 48, 24, and continuously reduce the learning rate of the optimizer algorithm with the training rounds; on this basis, optimize and adjust the model parameters to improve the accuracy of ozone forecast;

[0061] The backbone network of the coding layer is the key network for extracting time series features, and its performance plays an important role in the network's capture of time series features. The backbone network preset in the present invention is the coding network of the Transformer model, which accurately extracts features related to the time series, and provides effective information for the decoder to make sequence predictions.

[0062] C. Model training optimization

[0063] The present invention will train and optimize the model through the following process:

[0064] I) First, the ozone monitoring data set is divided into training set, validation set and test set. In order to test the applicability of the model, the months when severe ozone pollution occurs in Kunming, Yunnan Province (March to May) are selected as the test set of the model;

[0065] II) Since the units of meteorological data and air quality data are different and the magnitudes are quite different, in order to reduce the model training error, the present invention will use the normal distribution normalization method to eliminate the impact of the dimensional differences between pollutants and meteorological data;

[0066] III) The batch size is selected as 256, and the mini-batch gradient descent method is used to optimize the model training. In addition, to ensure the randomness of each batch data selection, the training set is shuffled when using the mini-batch gradient descent method;

[0067] IV) The model training judgment condition set in the present invention is: when the maximum number of training rounds is reached or the loss function value of the validation set does not decrease for 10 consecutive rounds, it is determined that the condition is met and the training is stopped to save the parameters; otherwise, the training is returned to the next round of training;

[0068] V) Finally, in order to scientifically evaluate the prediction accuracy of the model, the present invention selects the root mean square error (RMSE), absolute error (MAE) and determination coefficient (R 2 ) as the evaluation parameter of the model.

[0069] Step 3: Use the evaluation results of the optimized Yunnan ozone forecast model in the validation set as an indicator to evaluate the quality of the deep learning model and conduct ozone concentration evaluation;

[0070] The evaluation indicators include mean square error MSE, mean absolute error MAE, and determination coefficient R 2, select evaluation indicators

[0071] Metric=(MSE+MAE) / 2-R 2 As an evaluation indicator.

[0072] The model with the lowest metric on the validation set is selected as the optimal model and used for fast and accurate forecasting.

[0073] Example 2

[0074] Specific implementation steps of the Transformer model based on impact factor analysis and peak-aware optimization:

[0075] 1. Calculate channel features

[0076] First, the multivariate time series data of the input joint climate data and pollutant data are preprocessed and embedded; specifically:

[0077] The multivariate time series data of the combined climate data and pollutant data are denoted as Where: B represents the batch size; L represents the length of the input time series; num factors Represents the number of climate and pollutant types. The variables here are meteorological observation data, meteorological numerical simulation data, and pollutant observation data, specifically: historical O3, CO, SO2, NO2, PM 2.5 、PM 10、 Altitude, radiation, sea level pressure, vertical wind speed, temperature, wind speed, wind direction, total cloud cover, relative humidity and atmospheric pressure.

[0078] Map the data to a high-dimensional feature space through an embedding layer and calculate the channel features:

[0079] Feature channel =Embedding value (x enc )+Embedding position (x enc );

[0080] in, Embedding value Map the original numerical data x to a high-dimensional feature space to capture the numerical information of the data; Embedding position Captures position information of each position in the input sequence, helping the model understand the relative position of elements in the sequence.

[0081] 2. Extract global features of channels

[0082] 2.1 Perform average pooling on the features of each channel in the time dimension to obtain the global feature representation of each channel: MFeature channel =Mean(Feature channel , dim=-1) shape is

[0083] This process compresses the time series information into a fixed-length vector, thereby providing a simplified feature representation for subsequent impact factor analysis attention calculation.

[0084] 3. Conduct impact factor analysis and calculate impact scores

[0085] 3.1 Using the channel attention mechanism, calculate the attention score of each channel. The specific steps are:

[0086] Score channel =Attention channel (MFeature channel ),

[0087] Among them, the impact score is Here, Attention channel It consists of two fully connected layers. The first fully connected layer maps the input, then passes through the activation function layer (ReLU), then passes through the second fully connected layer, and finally uses Softmax for normalization to output the attention score of each channel. channel The attention score here is the impact score.

[0088] 3.2 To identify the quantitative impact of different factors on ozone forecasting, select Score channel The largest n impact scores have their corresponding impact factors taken as n-main controlling factors.

[0089] 4. Expand channel scores to match input dimensions

[0090] The channel scores are dimensionalized so that they are weighted with the input data.

[0091] ExpandedScore channel =Score channel .unsqueeze(1).repeat(1,L,1)

[0092] The impact score after expanding the dimension is Use unsqueeze(1) to add a dimension after the first dimension of the channel score; use repeat(1,L,1) to replicate the channel score in the time dimension to ensure that it matches the time dimension of the input data.

[0093] 5. Weight each channel of the input data

[0094] Weight each channel of the input data:

[0095] x enc_weighted =Encoder data (x enc ,x mark )×ExpandedScore channel ,

[0096] in, Embedding layer Encoder data (x enc ,x mark ) is calculated as:

[0097] Encoder data (x enc ,x mark )=Embedding value (x enc )+

[0098] Embedding position (x enc )+Embedding temporal (x mark );

[0099] Embedding value The original numerical data x enc Mapping to a high-dimensional feature space to capture the numerical information of the data.

[0100] Embedding temporal Extract time-related information such as timestamps, dates, or periodic features.

[0101] Embedding position Captures position information of each position in the input sequence, helping the model understand the relative position of elements in the sequence.

[0102] Multiply the weighted channel scores by the original input data element by element. Channels with high importance have a greater impact on the model due to their higher attention scores, while channels with lower attention scores have a smaller impact.

[0103] 6. Encoder part

[0104] 6.1. Embed the weighted input data and prepare it for the encoder:

[0105] ENC out =Encoder input (x enc ,x mark )

[0106] The Encoder input Network architecture and Encoder data Consistent, Enc out The shape is [B, L, D], x mark is the time stamp of the input sequence;

[0107] 6.2. Input the embedded data into the encoder to extract high-level feature representation:

[0108] TEnc out = TransformerEncoder(Enc out ,AttentionMask=None)

[0109] The encoder adopts the Transformer architecture. TransformerEncoder consists of two layers, which can effectively extract the context information of the input sequence. Each layer contains a multi-head self-attention mechanism and a feedforward neural network, and uses residual connections and layer normalization to enhance the training effect of the model.

[0110] 7. Decoder part

[0111] 7.1. Embed the decoder input data:

[0112] Dec out =Decoder input (x decode ,x markdec )

[0113] The Decoder input Network architecture and Encoder data Consistent, x decode is an all-zero sequence with dimension x markdec is the time mark of the prediction sequence, Dec out is the decoding feature Dec of the decoder out ∈R B×L‘×D . L' represents the length of the decoder sequence.

[0114] 7.2. Input the embedded data and the encoder output into the decoder for decoding

[0115] TDecout = TransformerDecoder(TEnc out )

[0116] Among them, TDec out ∈R B×L‘×D ,The decoder uses an autoregressive Transformer architecture to generate prediction sequences. Through the self-attention mechanism, the decoder can use the context information output by the encoder to generate the output of the current time step and guide the subsequent generation process through the historical output information.

[0117] 8. Output prediction results

[0118] 8.1. Convert the decoder output to the final prediction result through linear layers or other mappings

[0119] FinalOut = projection(TDec out ),

[0120] Among them, projection is a linear layer or other mapping function, which is used to map the output of the decoder to the target dimension to obtain the final prediction result FinalOut, FinalOut∈R B×L‘×target_dim .

[0121] In the above steps, B represents the batch size; L represents the length of the time series; num_factors represents the number of climate and pollutant types; D represents the feature space dimension; target_dim represents the dimension of the prediction target, which is 1 in the ozone prediction process.

[0122] 9. Peak Perceptual Loss Function

[0123] The peak perception loss function is used during the training process to improve the model's ability to perceive the peak. Specifically,

[0124] 9.1 Use the peak search algorithm to find the peak position, including:

[0125] PeakIndex=PeakDetection(Target)

[0126] Where Target is the true value of the predicted ozone concentration, Target∈R B×L‘ , PeakIndex is the index of the peak, PeakIndex∈{1,2,3,...,L'}; PeakDetection is the peak search algorithm, here is the Automatic multiscale-based peak detection (AMPD) algorithm;

[0127] 9.2 Calculate the peak perceptual loss function and perform gradient backpropagation

[0128] Loss peak =MSELoss(Target PeakIndex ,FinalOut PeakIndex )

[0129] Target PeakIndex , indicating that by Target i The sequence of numbers, where i∈PeakIndex. Similarly FinalOut PeakIndex , indicating that FinalOut i The sequence of numbers, where i∈PeakIndex;

[0130] Loss MSE =MSELoss(Target,FinalOut)

[0131] Loss total =Loss MSE +aLoss peak

[0132] The α in the formula is set to 1.

[0133] In summary, by increasing the penalty level of the model at the corresponding peak value, the model can better perceive the peak time and the corresponding concentration, improve the ozone peak forecast effect, and achieve rapid and accurate forecast of future severe ozone pollution processes in Yunnan Province; thus, the optimization of the ozone forecast model in Yunnan Province can be completed.

Claims

1. A method for predicting ozone concentration based on peak perception and influencing factor analysis, characterized in that: The following steps are involved: S1. Obtain pollutant and meteorological monitoring data in Yunnan for many years through network data collection, and perform outlier processing, null value filling, data normalization and data set division on the pollutant and meteorological monitoring data to build a large database of pollutant and meteorological data in Yunnan that supports deep learning; S2. The Transformer model based on influencing factor analysis and peak perception optimization is used to train the historical observation data set in the pollutant and meteorological data database in Yunnan, and the ozone forecast model in Yunnan based on historical meteorology, pollutants and future ozone is established. Based on big data deep learning, the ozone forecast model in Yunnan is optimized through peak perception and influencing factor analysis, and the n-master control factors are output; S3. The evaluation results of the optimized Yunnan ozone forecast model in the validation set are used as indicators to evaluate the quality of the deep learning model and to evaluate the ozone concentration. The evaluation indicators include MSE, MAE and R 2 , the evaluation index expression is selected as follows: Metric=(MSE+MAE) / 2-R 2 ; The model with the lowest metric on the validation set is selected as the optimal model and used for fast and accurate forecasting; MSE is mean square error, MAE is absolute error, R 2 is the determination coefficient.

2. The ozone concentration prediction method based on peak perception and influencing factor analysis according to claim 1 is characterized in that: In S1, the large database of pollutants and meteorological data in Yunnan is established as follows: Through online data download, the six conventional monitoring pollutants (O3, CO, SO2, NO2, PM 2.5 、PM 10 ) Continuous online observation data and meteorological observation data are combined with ECMWF meteorological forecast products to provide hourly-scale meteorological numerical simulation data on boundary layer height, radiation, sea level pressure, and vertical wind speed to collect and organize, and build a long-term historical data set of ozone, meteorology, and other pollutants; and perform data outlier processing, null value filling, data normalization, and data set division on the data set to meet the data format required for deep learning, and establish a large database of pollutants and meteorological data in Yunnan to support deep learning; During the project implementation period, meteorological and pollutant observation data will be continuously stored and sorted to achieve dynamic updating of the large database.

3. The ozone concentration prediction method based on peak perception and influencing factor analysis according to claim 2 is characterized in that: The data outlier processing, null value filling, data normalization and data set division are as follows: Data outlier processing: abnormal reasons such as equipment damage may cause the monitoring data to be too large or the equipment to be insensitive, resulting in the monitoring data being continuously 0 for a period of time. This part of the data content in the conventional monitoring pollutant data and meteorological data is eliminated; Filling in missing values: Due to observation equipment or observation conditions, the phenomenon of discontinuity of observation data is obvious, so it is very important to supplement missing values. The data filling method used in this project is based on the interpolation of statistical values ​​at different time scales: V ij =|V monthij +V weekdayij +V hourij -2×V yearij | Where V ij is the filling value of pollutant i at missing time j, Vmonth ij is the mean concentration of pollutant i in the month of missing time j, Vweekday ij is the mean concentration of pollutant i in the week of missing time j, Vyeay ij is the mean pollutant concentration of pollutant i in the year of missing time j; Data normalization: Data normalization is conducive to improving the stability of training and the convergence speed of model training. This project adopts the normal distribution normalization method, which is to subtract the mean of the characteristic attributes of the data and divide it by the variance; convert it into a standard normal distribution with a mean of 0 and a variance of 1: x=(x-μ) / σ In the formula, μ is the data mean, σ is the data variance. After standardizing the training set by recording the mean and variance on the training set, when the model is used in subsequent tests, the test data is standardized using the variance and mean on the training set. Dataset division: The data used are from more than 100 ground observation stations in the Yunnan monitoring network that began to monitor O3, NO2, CO, SO2, PM2.5 and 3.5% of the total CO2 emissions in 2015. 2.5 , PM 10 Observe data continuously hourly; select data from the past year as the test set to evaluate the model capability, and divide the other data into training set and validation set according to the year in a 4:1 ratio; the training set is used to train the model parameters, and the validation set is used to select the model with the best number of training rounds.

4. The ozone concentration prediction method based on peak perception and influencing factor analysis according to claim 1 is characterized in that: In S2, based on big data deep learning, the ozone forecast model in Yunnan is optimized through peak perception and influencing factor analysis as follows: S2.

1. The Transformer model based on influencing factor analysis is used to train the historical data set. The meteorological and atmospheric pollutant monitoring data of more than ten years and the numerical simulation data provided by the ECMWF meteorological forecast products are used as the input of the neural network model, and the predicted ozone concentration is used as the output. The ozone forecast model for the Yunnan region at the hourly scale and the daily scale is established respectively; and the perception score is output, and the largest n perception scores are selected, and the corresponding influencing factors are taken as n-main control factors; S2.

2. Simultaneously apply the peak sensing method to find the corresponding peak values ​​in the daily forecast and hourly forecast results using the peak finding algorithm, including: PeakIndex=PeakDetection(Target) Where Target is the true value of the predicted ozone concentration, Target∈R B×L‘ , PeakIndex is the index of the peak, PeakIndex∈{1,2,3,...,L'}; PeakDetection is the peak search algorithm, here is the Automatic multiscale-based peak detection (AMPD) algorithm; Calculate the peak perceptual loss function and perform gradient backpropagation Loss peak =MSELoss(Target PeakIndex ,FinalOUt PeakIndex ) Target PeakIndex , indicating that by Target i The sequence of numbers, where i∈PeakIndex; similarly FinalOut PeakIndex , indicating that FinalOut i The sequence of numbers, where i∈PeakIndex; Loss MSE =MSELoss(Target,FinalOut) Loss total =Loss MSE +αLoss peak In the formula, α is 1; Then, the penalty level of the model is increased at the corresponding peak value, so that the model can better perceive the peak time and the corresponding concentration, improve the ozone peak forecast effect, and realize the rapid and accurate forecast of future severe ozone pollution process in Yunnan Province; thus, the optimization of the ozone forecast model in Yunnan Province can be completed.

5. The ozone concentration prediction method based on peak perception and influencing factor analysis according to claim 1 is characterized in that: The meteorological data set includes temperature, wind speed, wind direction, total cloud cover, relative humidity, atmospheric pressure, boundary layer height, radiation, sea level pressure, and vertical wind speed.

Images