Ozone over-standard prediction method and device, storage medium and electronic equipment

Abstract

The present application provides an ozone over-standard prediction method, device, storage medium and electronic equipment, the method comprising: obtaining an ozone over-standard training data set in a target area; based on the ozone over-standard training data set, respectively constructing an ozone over-standard prediction model under each undetermined meteorological parameter combination in a plurality of undetermined meteorological parameter combinations; and respectively determining a prediction performance index value of the ozone over-standard prediction model under each undetermined meteorological parameter combination; based on the prediction performance index value of the ozone over-standard prediction model under each undetermined meteorological parameter combination, determining a key meteorological parameter combination, and taking the ozone over-standard prediction model under the key meteorological parameter combination as a target ozone over-standard prediction model of the target area, the target ozone over-standard prediction model being used for ozone over-standard prediction according to the key meteorological parameter combination. The present application embodiment can conveniently perform ozone over-standard prediction and improve the accuracy of ozone over-standard prediction.

Description

Technical Field

[0001] This invention relates to the field of air quality prediction technology, and in particular to a method, device, storage medium, and electronic equipment for predicting ozone exceedances. Background Technology

[0002] Currently, near-surface ozone (O3) pollution has become increasingly prominent in recent years and has received widespread attention. However, related technologies typically use numerical air quality models to simulate ozone concentrations and then use these concentrations to determine whether ozone levels exceed standards. However, the area near the secondary limit for ozone concentration (160 μg / m³) can introduce a critical error in determining whether ozone levels exceed standards, leading to complex calculations and low accuracy in ozone exceedance predictions. Therefore, there is currently no satisfactory solution for conveniently predicting ozone exceedances and improving their accuracy. Summary of the Invention

[0003] In view of this, embodiments of the present invention provide a method, apparatus, storage medium, and electronic device for predicting ozone exceedances, to solve the problems of complex calculations and low accuracy of ozone exceedance prediction in related technologies; that is, embodiments of the present invention can conveniently predict ozone exceedances, effectively reduce the calculation cost and time consumption of ozone exceedance prediction, and can effectively improve the accuracy of ozone exceedance prediction by using a target ozone exceedance prediction model.

[0004] According to one aspect of the present invention, a method for predicting ozone exceedance is provided, the method comprising:

[0005] Obtain a training dataset of ozone exceedances within the target area. Each ozone exceedance training dataset includes an ozone exceedance training label and a meteorological parameter dataset. Each meteorological parameter dataset includes the parameter values ​​of each meteorological parameter among multiple meteorological parameters.

[0006] Based on the ozone exceedance training dataset, ozone exceedance prediction models are constructed for each of the multiple combinations of undetermined meteorological parameters; and the prediction performance index values ​​of the ozone exceedance prediction models for each combination of undetermined meteorological parameters are determined.

[0007] Based on the prediction performance index values ​​of the ozone exceedance prediction model under each combination of undetermined meteorological parameters, key meteorological parameter combinations are determined, and the ozone exceedance prediction model under the key meteorological parameter combinations is used as the target ozone exceedance prediction model for the target area. The target ozone exceedance prediction model is used to predict ozone exceedance according to the key meteorological parameter combinations.

[0008] According to another aspect of the present invention, an ozone exceedance prediction device is provided, the device comprising:

[0009] The acquisition unit is used to acquire a set of ozone exceedance training data in the target area. One ozone exceedance training data includes an ozone exceedance training label and a meteorological parameter data. One meteorological parameter data includes the parameter values ​​of each meteorological parameter among multiple meteorological parameters.

[0010] The processing unit is used to construct ozone exceedance prediction models for each of the multiple combinations of undetermined meteorological parameters based on the ozone exceedance training data set; and to determine the prediction performance index values ​​of the ozone exceedance prediction models for each combination of undetermined meteorological parameters.

[0011] The processing unit is further configured to determine key meteorological parameter combinations based on the prediction performance index values ​​of the ozone exceedance prediction models under each combination of undetermined meteorological parameters, and to use the ozone exceedance prediction models under the key meteorological parameter combinations as the target ozone exceedance prediction models for the target area. The target ozone exceedance prediction models are used to predict ozone exceedances according to the key meteorological parameter combinations.

[0012] According to another aspect of the present invention, an electronic device is provided, the electronic device including a processor and a memory storing a program, wherein the program includes instructions that, when executed by the processor, cause the processor to perform the methods mentioned above.

[0013] According to another aspect of the present invention, a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the methods mentioned above is provided.

[0014] This invention provides an embodiment of ozone exceedance training data set within a target area. Each ozone exceedance training data set includes an ozone exceedance training label and meteorological parameter data, with each meteorological parameter data set including the parameter values ​​of various meteorological parameters. Then, based on the ozone exceedance training data set, ozone exceedance prediction models can be constructed for each of the multiple combinations of undetermined meteorological parameters; and the prediction performance index value of each ozone exceedance prediction model under each combination of undetermined meteorological parameters can be determined. Furthermore, based on the prediction performance index values ​​of the ozone exceedance prediction models under each combination of undetermined meteorological parameters, key meteorological parameter combinations can be determined, and the ozone exceedance prediction model under the key meteorological parameter combinations can be used as the target ozone exceedance prediction model for the target area. The target ozone exceedance prediction model is used to predict ozone exceedances according to the key meteorological parameter combinations. As can be seen, the embodiments of the present invention can conveniently predict ozone exceedances, effectively reduce the computational cost and time consumption of ozone exceedance prediction; furthermore, the embodiments of the present invention can determine the combination of key meteorological parameters to determine the target ozone exceedance prediction model for the target area, thereby enabling ozone exceedance prediction to be performed according to the combination of key meteorological parameters using the target ozone exceedance prediction model, which can effectively improve the accuracy of ozone exceedance prediction, thus improving the accuracy of ozone exceedance prediction for the target area. Attached Figure Description

[0015] Further details, features, and advantages of the invention are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which:

[0016] Figure 1 A flowchart illustrating an ozone exceedance prediction method according to an exemplary embodiment of the present invention is shown.

[0017] Figure 2 A flowchart illustrating another method for predicting ozone exceedances according to an exemplary embodiment of the present invention is shown;

[0018] Figure 3 A flowchart illustrating another method for predicting ozone exceedance according to an exemplary embodiment of the present invention is shown;

[0019] Figure 4 A schematic diagram of the AUC value of an ozone exceedance prediction model under a combination of undetermined meteorological parameters according to an exemplary embodiment of the present invention is shown.

[0020] Figure 5 A schematic diagram illustrating the relationship between AUC value and the number of meteorological parameters according to an exemplary embodiment of the present invention is shown.

[0021] Figure 6 A schematic block diagram of an ozone exceedance prediction device according to an exemplary embodiment of the present invention is shown;

[0022] Figure 7 A structural block diagram of an exemplary electronic device that can be used to implement embodiments of the present invention is shown. Detailed Implementation

[0023] Embodiments of the present invention will now be described in more detail with reference to the accompanying drawings. While some embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the invention. It should be understood that the accompanying drawings and embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the invention.

[0024] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0025] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first", "second", etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0026] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0027] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0028] It should be noted that the execution subject of the ozone exceedance prediction method provided in this embodiment of the invention can be one or more electronic devices, and this invention does not limit this; wherein, the electronic device can be a terminal (i.e., a client) or a server. Therefore, when the execution subject includes multiple electronic devices, and among the multiple electronic devices includes at least one terminal and at least one server, the ozone exceedance prediction method provided in this embodiment of the invention can be executed jointly by the terminal and the server. Accordingly, the terminal mentioned herein can include, but is not limited to: smartphones, tablets, laptops, desktop computers, etc.; the server mentioned herein can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms, etc.

[0029] Based on the above description, this invention proposes an ozone exceedance prediction method, which can be executed by the aforementioned electronic device (terminal or server); or, the ozone exceedance prediction method can be executed jointly by the terminal and the server. For ease of explanation, the following description will use the execution of the ozone exceedance prediction method by an electronic device as an example; such as Figure 1 As shown, the ozone exceedance prediction method may include the following steps S101-S103:

[0030] S101, Obtain a set of ozone exceedance training data in the target area. One ozone exceedance training data includes an ozone exceedance training label and a meteorological parameter data. One meteorological parameter data includes the parameter values ​​of each meteorological parameter among multiple meteorological parameters.

[0031] In this embodiment of the invention, an ozone exceedance training tag in an ozone exceedance training dataset can be used to indicate whether ozone exceeds the standard within the time period corresponding to the ozone exceedance training dataset. That is, a tag (also called an ozone exceedance status tag, such as an ozone exceedance training tag or an ozone exceedance verification tag as described below) can be used to indicate whether ozone exceeds the standard within the time period corresponding to the data. The time corresponding to a dataset (such as ozone exceedance training data, etc.) can refer to the time in which the data is located. For example, if an ozone exceedance training dataset includes meteorological parameter data and an ozone exceedance training tag within March 5, 2022, then the time corresponding to the ozone exceedance training dataset is March 5, 2022. In other words, an ozone exceedance training dataset within a target area can include an ozone exceedance training tag and meteorological parameter data within a time period for the target area, and so on. Optionally, a label can be an ozone exceedance indicator or an ozone not exceedance indicator. The ozone exceedance indicator can be used to indicate that ozone exceeds the standard, and the ozone not exceedance indicator can be used to indicate that ozone does not exceed the standard. Alternatively, it can be an ozone exceedance probability vector (which may include the probability of ozone exceeding the standard and the probability of ozone not exceeding the standard), such as (1,0) which can represent that the probability of ozone exceeding the standard is 1 and the probability of ozone not exceeding the standard is 0, etc. This embodiment of the invention does not limit this. Optionally, both the ozone exceedance indicator and the ozone not exceedance indicator can be set according to experience or according to actual needs. This embodiment of the invention does not limit this. For example, 1 can represent ozone exceeding the standard, 0 can represent ozone not exceeding the standard, etc. Based on this, a label can be used to indicate the probability of ozone exceeding the standard. For example, if a label indicates that ozone exceeds the standard, the corresponding probability of ozone exceeding the standard can be 1, and if a label indicates that ozone does not exceed the standard, the corresponding probability of ozone exceeding the standard can be 0. Optionally, a data set (such as ozone exceedance training data or ozone exceedance verification data) may also include the corresponding time period.

[0032] Optionally, the aforementioned meteorological parameters may include, but are not limited to: temperature, specific humidity, cloud cover, total precipitation, pressure, horizontal wind speed, vertical wind speed, atmospheric boundary layer height, ozone column concentration, etc., and this embodiment of the invention does not limit these parameters. Optionally, the target area may be any area (such as any city), and this embodiment of the invention does not limit this area. Based on this, this embodiment of the invention can construct corresponding target ozone exceedance prediction models for different areas (such as different cities) to predict ozone exceedances. In other words, this embodiment of the invention can achieve ozone exceedance prediction at a regional scale (such as a city scale). Optionally, an ozone exceedance training dataset may also be called an ozone exceedance determination training dataset, etc.

[0033] Optionally, the training dataset of ozone exceedances in the target area can be obtained in the following ways, including but not limited to:

[0034] The first acquisition method: The electronic device stores ozone exceedance training data sets in multiple regions in its own storage space, and these multiple regions include the target region. In this case, the electronic device can obtain the ozone exceedance training data sets in the target region from its own storage space.

[0035] The second method is to use electronic devices to obtain training data download links and download ozone exceedance training data sets within the target area using these links.

[0036] The third acquisition method: Electronic devices can acquire ozone exceedance labels and meteorological parameter data for each time period (also known as time intervals or time ranges) within a target area, and perform data processing and / or data segmentation on the acquired data to obtain a training dataset of ozone exceedance data within the target area, etc. Optionally, multiple times can include any time period, and this embodiment of the invention does not limit this; for example, multiple times can include each day from 2015 to 2022, thus collecting daily ozone exceedance labels and meteorological parameter data for the target area from 2015 to 2022, etc. Optionally, a single time period can be a single day, etc., and this embodiment of the invention does not limit this.

[0037] Optionally, the electronic device can also collect ozone concentration data for each time period within a target area, to determine an ozone exceedance label for the target area at the corresponding time period based on the ozone concentration data for each time period (such as an ozone exceedance training label). Optionally, the ozone concentration data can come from sources such as national monitoring stations. For example, it can collect the daily maximum 8-hour average ozone concentration data (which can be represented as MDA8) from at least one air quality monitoring station in the target area, and perform an average calculation on the daily maximum 8-hour average ozone concentration data from at least one air quality monitoring station in the target area to obtain the daily maximum 8-hour average ozone concentration data for the target area (i.e., the ozone concentration in the target area within a certain time period). For example, for any given time period, the ozone exceedance status of the target area at any given time can be classified based on my country's secondary concentration limits for ambient air pollutants. If the MDA8 of the target area at any given time is less than or equal to 160 μg / m³, then the ozone level in the target area at any given time is determined to be within the limit, and the "ozone not exceeding the limit" label is used as the label for the ozone exceedance status of the target area at any given time. Otherwise, the ozone level in the target area at any given time is determined to be within the limit, and the "ozone exceeding the limit" label is used as the label for the ozone exceedance status of the target area at any given time. Optionally, the ozone exceedance status (also referred to as the ozone exceedance category) can be either "ozone exceedance" or "ozone not exceedance". Optionally, the meteorological parameter data of the target area at any given time can come from the NASA GEOS-FP grid-level meteorological parameter database (a gridded meteorological parameter database with a latitude resolution of 0.25° and a longitude resolution of 0.3125°), etc., and this embodiment of the invention does not limit this. Optionally, the electronic device can distribute grid-level meteorological data to various regions (such as cities) based on latitude and longitude correspondence, thereby obtaining the average value of meteorological parameter data of each grid in the target region at any given time, which can then be used as the meteorological parameter data of the target region at any given time. For example, for any meteorological parameter among multiple meteorological parameters, the parameter value of any meteorological parameter in each grid in the target region at any given time can be determined, and the average value of any meteorological parameter in each grid in the target region at any given time can be calculated to obtain the parameter value of any meteorological parameter in the target region at any given time. The meteorological parameter data of the target region at any given time may include the parameter values ​​of all meteorological parameters in the target region at any given time. Optionally, a region may include all grids whose grid center points are located in the corresponding region, that is, the region where the grid center point of a grid is located can be used as the region where the corresponding grid is located, and so on. Optionally, if the acquired data has missing data (such as not including the parameter value of a certain meteorological parameter at a certain time), the missing values ​​can also be processed (such as interpolation processing, etc.), and so on.

[0038] Optionally, the electronic device can divide the ozone exceedance label and meteorological parameter data of the target area at each time point across multiple time periods into an ozone exceedance training dataset (e.g., including data from 2015-2021) and an ozone exceedance validation dataset (e.g., including data from 2022). Based on this, the ozone exceedance training dataset of the target area can be obtained. Alternatively, the data can be divided into a first part of the data within the target area (e.g., including data from 2015-2021, which can be used to screen key meteorological parameters) and a second part of the data (e.g., including data from 2022, which can be used to test the robustness of the model's predictive performance). The first part of the data is randomly shuffled and divided into P parts. Each of the P-1 parts can be used as an ozone exceedance training dataset within the target area, the remaining part of the P parts can be used as an ozone exceedance validation dataset within the target area, and the second part of the data can be used as an ozone exceedance test dataset within the target area, where P is a positive integer. Based on this, when P is greater than 2, the number of ozone exceedance training datasets within the target area can be multiple. For example, such as Figure 2 As shown, taking one day as an example, assuming P is 5, the first part of the data includes ozone and meteorological data from 2015 to 2021, and the second part of the data includes data from 2022. Then, the first part of the data can be randomly shuffled to divide it into 4 ozone exceedance training data sets (referred to as training sets) and 1 ozone exceedance validation data set (referred to as validation sets) to determine the key meteorological parameters (i.e., the meteorological parameters in the combination of key meteorological parameters). The second part of the data can be used as the ozone exceedance test data set (referred to as test set) to verify the robustness of the model, and so on.

[0039] S102, based on the ozone exceedance training dataset, construct ozone exceedance prediction models for each of the multiple unknown meteorological parameter combinations; and determine the prediction performance index values ​​of the ozone exceedance prediction models for each unknown meteorological parameter combination.

[0040] Optionally, when there are multiple ozone exceedance training datasets within the target area, for any one of the multiple unknown meteorological parameter combinations, the electronic device can construct multiple initial ozone exceedance prediction models under any unknown meteorological parameter combination based on each ozone exceedance training dataset in the multiple unknown meteorological parameter combinations. Based on the prediction performance index values ​​of each initial ozone exceedance prediction model under any unknown meteorological parameter combination, an ozone exceedance prediction model for any unknown meteorological parameter combination is determined from among the multiple initial ozone exceedance prediction models under any unknown meteorological parameter combination. The ozone exceedance prediction model for any unknown meteorological parameter combination can be the initial ozone exceedance prediction model with the best prediction performance among the multiple initial ozone exceedance prediction models under any unknown meteorological parameter combination. Specifically, one of the multiple ozone exceedance training datasets can be used to construct one initial ozone exceedance prediction model for any unknown meteorological parameter combination. It should be noted that the implementation methods, such as constructing an initial ozone exceedance prediction model for any combination of undetermined meteorological parameters based on an ozone exceedance training dataset, and determining the prediction performance index values ​​of each initial ozone exceedance prediction model for any combination of undetermined meteorological parameters, are the same as the implementation methods, such as constructing ozone exceedance prediction models for each combination of undetermined meteorological parameters based on an ozone exceedance training dataset, and determining the prediction performance index values ​​of each ozone exceedance prediction model for each combination of undetermined meteorological parameters. These details will not be repeated here. Optionally, an ozone exceedance prediction model can be a Logistic Generalized Additive Model (LGAM), or a neural network model, etc., and this embodiment of the invention does not limit this.

[0041] Optionally, the prediction performance index value of an ozone exceedance prediction model can be used to indicate the prediction performance of the corresponding ozone exceedance prediction model; optionally, a prediction performance index value can be an AUC value (an index used to measure the performance of a binary classification model, the value of AUC is usually between 0.5 and 1, the closer to 1, the better the prediction performance, and when AUC=0.5, it means there is no prediction performance, that is, the larger the prediction performance index value, the better the prediction performance of the ozone exceedance prediction model), or a Brier-score value (an index used to evaluate the prediction accuracy of a classification model, the Brier-score is a number in the range of 0-1, the closer to 0, the smaller the difference between the prediction result and the actual result, that is, the smaller the prediction performance index value, the better the prediction performance of the ozone exceedance prediction model), etc., and the embodiments of the present invention do not limit this.

[0042] S103. Based on the prediction performance index values ​​of the ozone exceedance prediction model under each combination of undetermined meteorological parameters, determine the key meteorological parameter combination, and use the ozone exceedance prediction model under the key meteorological parameter combination as the target ozone exceedance prediction model for the target area. The target ozone exceedance prediction model is used to predict ozone exceedance according to the key meteorological parameter combination.

[0043] Based on this, when a time period is a day, the embodiments of the present invention can use the target ozone exceedance prediction model to predict ozone exceedance on a daily time scale within the target area. In practical applications, the target ozone prediction model can be invoked to predict the ozone exceedance situation in the target area within a target time period according to the combination of key meteorological parameters. Thus, ozone exceedance prediction can be performed for the target area and target time according to the combination of key meteorological parameters, and the ozone exceedance prediction result of the target area within the target time period can be obtained.

[0044] Optionally, an ozone exceedance prediction result can be an ozone exceedance indicator or an ozone not exceedance indicator, or it can be an ozone exceedance prediction probability value (used to indicate the probability of ozone exceeding the standard, such as a value greater than 0.5 indicating ozone exceedance, and a value less than or equal to 0.5 indicating ozone not exceeding the standard), etc., and this embodiment of the invention does not limit this. Optionally, the target time can be any time, such as any future day, and this embodiment of the invention does not limit this. Optionally, when an ozone exceedance prediction result is an ozone exceedance indicator or an ozone not exceedance indicator, after performing ozone exceedance prediction for the target area and target time according to the combination of key meteorological parameters, and obtaining the ozone exceedance prediction probability value for the target area within the target time, the ozone exceedance prediction result for the target area within the target time can be determined based on the ozone exceedance prediction probability value. For example, if the ozone exceedance prediction probability value is greater than 0.5, the ozone exceedance indicator can be used as the ozone exceedance prediction result for the target area within the target time; if the ozone exceedance prediction probability value is less than or equal to 0.5, the ozone not exceedance indicator can be used as the ozone exceedance prediction result for the target area within the target time, and so on.

[0045] Optionally, when predicting ozone exceedances for a target area and a target time, the parameter values ​​of each key meteorological parameter in the key meteorological parameter group for the target area within the target time can be obtained. Based on these values, ozone prediction data can be determined, and the target ozone exceedance prediction model can be invoked. Ozone exceedance prediction can then be performed based on the ozone prediction data to obtain the ozone exceedance prediction result for the target area within the target time. Optionally, the ozone prediction data may include, but is not limited to, the parameter values ​​of each key meteorological parameter for the target area within the target time and / or the seasonal representation information of the season to which the target time belongs, etc. Optionally, the seasonal representation information of a season can be a value of a seasonal variable (also called a variable value); optionally, the seasonal representation information of a season can be set according to experience or actual needs, and this embodiment of the invention does not limit this; for example, the seasonal representation information for spring can be 0, the seasonal representation information for summer can be 1, the seasonal representation information for autumn can be 2, the seasonal representation information for winter can be 3, etc.

[0046] This invention provides an embodiment of ozone exceedance training data set within a target area. Each ozone exceedance training data set includes an ozone exceedance training label and meteorological parameter data, with each meteorological parameter data set including the parameter values ​​of various meteorological parameters. Then, based on the ozone exceedance training data set, ozone exceedance prediction models can be constructed for each of the multiple combinations of undetermined meteorological parameters; and the prediction performance index value of each ozone exceedance prediction model under each combination of undetermined meteorological parameters can be determined. Furthermore, based on the prediction performance index values ​​of the ozone exceedance prediction models under each combination of undetermined meteorological parameters, key meteorological parameter combinations can be determined, and the ozone exceedance prediction model under the key meteorological parameter combinations can be used as the target ozone exceedance prediction model for the target area. The target ozone exceedance prediction model is used to predict ozone exceedances according to the key meteorological parameter combinations. As can be seen, the embodiments of the present invention can conveniently predict ozone exceedances, effectively reduce the computational cost and time consumption of ozone exceedance prediction; furthermore, the embodiments of the present invention can determine the combination of key meteorological parameters to determine the target ozone exceedance prediction model for the target area, thereby enabling ozone exceedance prediction to be performed according to the combination of key meteorological parameters using the target ozone exceedance prediction model, which can effectively improve the accuracy of ozone exceedance prediction, thus improving the accuracy of ozone exceedance prediction for the target area.

[0047] Based on the above description, this embodiment of the invention also proposes a more specific method for predicting ozone exceedances. Accordingly, this ozone exceedance prediction method can be executed by the aforementioned electronic device (terminal or server); or, the ozone exceedance prediction method can be executed jointly by the terminal and the server. For ease of explanation, the following description will use the execution of this ozone exceedance prediction method by an electronic device as an example; please refer to [link to relevant documentation]. Figure 3The ozone exceedance prediction method may include the following steps S301-S305:

[0048] S301, Obtain a set of ozone exceedance training data in the target area. One ozone exceedance training data includes an ozone exceedance training label and a meteorological parameter data. One meteorological parameter data includes the parameter values ​​of each meteorological parameter among multiple meteorological parameters.

[0049] S302, determine the initial meteorological quantities.

[0050] Optionally, the initial meteorological data can be set based on experience or according to actual needs; this embodiment of the invention does not limit this.

[0051] S303, combine multiple meteorological parameters according to the initial meteorological quantity to obtain multiple combinations of undetermined meteorological parameters. The number of meteorological parameters in a combination of undetermined meteorological parameters is equal to the initial meteorological quantity.

[0052] For example, assuming the initial meteorological quantity is 2, and the multiple meteorological parameters include 9 parameters: temperature, specific humidity, cloud cover, total precipitation, pressure, horizontal wind speed, vertical wind speed, atmospheric boundary layer height, and ozone column concentration, then the number of unknown meteorological parameter combinations can be C9. 2 The combinations of multiple undetermined meteorological parameters are shown in Table 1:

[0053] Table 1

[0054]

[0055] It can be seen that the number of undetermined meteorological parameter combinations in the multiple undetermined meteorological parameter combinations at this time can be 36.

[0056] S304. Based on the ozone exceedance training dataset, ozone exceedance prediction models are constructed for each of the multiple combinations of undetermined meteorological parameters; and the prediction performance index values ​​of the ozone exceedance prediction models for each combination of undetermined meteorological parameters are determined.

[0057] Optionally, when constructing ozone exceedance prediction models for each of multiple undetermined meteorological parameter combinations based on the ozone exceedance training dataset, the electronic device can iterate through each of the multiple undetermined meteorological parameter combinations and use the currently iterated undetermined meteorological parameter combination as the current undetermined meteorological parameter combination. Then, based on the ozone exceedance training dataset, a parameter combination training dataset for the current undetermined meteorological parameter combination can be determined. The parameter combination training dataset for the current undetermined meteorological parameter combination may include an ozone exceedance training label and the variable value of each of the multiple independent variables in the current undetermined meteorological parameter combination. The multiple independent variables in the current undetermined meteorological parameter combination may include each meteorological parameter and / or seasonal variable in the current undetermined meteorological parameter combination. Based on this, the parameter combination training dataset for the current undetermined meteorological parameter combination can be used to construct an ozone exceedance prediction model for the current undetermined meteorological parameter group. After iterating through each of the multiple undetermined meteorological parameter combinations, the ozone exceedance prediction model for each undetermined meteorological parameter combination is obtained. When multiple independent variables include seasonal variables, the embodiments of the present invention can fully consider the different impacts of meteorological data on ozone in different seasons, thereby analyzing the relationship between ozone exceedance and meteorological conditions in more detail, which can effectively improve the predictive performance of the ozone exceedance prediction model.

[0058] Optionally, an ozone exceedance prediction model can be a logistic generalized additive model, and the ozone exceedance training label in an ozone exceedance training dataset can be used to indicate whether the ozone exceeds the standard within the corresponding time period of the ozone exceedance training dataset. Based on this, when constructing an ozone exceedance prediction model for the current unknown meteorological parameter combination using a parameter combination training dataset, the electronic device can determine the ozone exceedance prediction model to be fitted under the current unknown meteorological parameter combination. This ozone exceedance prediction model to be fitted may include a non-parametric smoothing function for each meteorological parameter in the current unknown meteorological parameter combination; and the ozone exceedance prediction model to be fitted can be fitted using the parameter combination training dataset under the current unknown meteorological parameter combination to obtain the ozone exceedance prediction model under the current unknown meteorological parameter combination. For example, the ozone exceedance prediction model to be fitted under the current unknown meteorological parameter combination can be as shown in Formula 1.1:

[0059] Formula 1.1

[0060] Where P(y=1|X) represents the probability of ozone exceeding the standard, and X j f can represent the j-th meteorological parameter (which can be a continuous variable) in the current combination of undetermined meteorological parameters. jLet represent the nonparametric smoothing function of the j-th meteorological parameter, 'season' represent the seasonal variable, 'γ' represent the coefficient of the seasonal variable, 'β0' represent the intercept, 'ε' represent the residual, and 'n' represent the number of meteorological parameters in the current combination of meteorological parameters to be determined. Based on this, after determining the intercept, the nonparametric smoothing function of each meteorological parameter in the current combination of meteorological parameters to be fitted, the coefficient of the seasonal variable, and the residual, the ozone exceedance prediction model under the current combination of meteorological parameters can be obtained. It can be seen that the embodiments of the present invention can fully consider the influence of the current combination of meteorological parameters and the seasonal variable on the probability of ozone exceedance, and the ozone exceedance prediction model can be fitted by the nonparametric smoothing function of each meteorological parameter. The nonparametric smoothing function can be any form of function, thus fitting the data in a more flexible way, making the fitted ozone exceedance prediction model more accurate, thereby improving the prediction performance of the ozone exceedance prediction model. Optionally, a nonparametric smoothing function can be a spline function (in which case the order and other parameters of the spline function can be fitted), etc., and the embodiments of the present invention do not limit this. It should be noted that the specific implementation method for fitting the ozone exceedance prediction model in this embodiment of the invention is not limited. For example, the electronic device can use the parameter combination training data set under the current unknown meteorological parameter combination to model each meteorological parameter and dependent variable (i.e., the probability of ozone exceedance) under the current unknown meteorological parameter combination separately to determine the nonparametric function of each meteorological parameter, and then add them together to obtain a logical generalized additive model (at this time, the ozone exceedance prediction model to be fitted also includes parameters to be fitted, such as intercept, etc.), and finally calculate the parameters to be fitted in the ozone exceedance prediction model to be fitted; or, it can first model any meteorological parameter and dependent variable under the current unknown meteorological parameter combination separately to determine the nonparametric function of any meteorological parameter, and then model an undetermined meteorological parameter (i.e., a meteorological data whose nonparametric function is not determined) and dependent variable based on the function of the determined meteorological parameter, until the nonparametric function of each meteorological parameter is determined, and then fit the remaining parameters to be fitted; or, it can also call an open-source generalized additive model fitting tool to perform the fitting, etc.

[0061] Optionally, when determining the parameter combination training data set under the current undetermined meteorological parameter combination based on the ozone exceedance training data set, for any ozone exceedance training data in the ozone exceedance training data set, the electronic device can determine the parameter combination training data under the current undetermined meteorological parameter combination based on any ozone exceedance training data. This means determining the ozone exceedance training label and the parameter value (also called variable value) of each meteorological parameter in the current undetermined meteorological parameter combination from any ozone exceedance training data, and adding it to the parameter combination training data under the current undetermined meteorological parameter combination for any ozone exceedance training data. And / or the value of a seasonal variable can be determined based on the time corresponding to any ozone exceedance training data (that is, the seasonal representation information of the season to which the time corresponding to any ozone exceedance training data belongs can be used as the value of the seasonal variable) and added to the parameter combination training data of any ozone exceedance training data under the current undetermined meteorological parameter combination. This achieves the determination of the parameter combination training data of any ozone exceedance training data under the current undetermined meteorological parameter combination. The parameter combination training data set under the current undetermined meteorological parameter combination may include the parameter combination training data of each ozone exceedance training data set under the current undetermined meteorological parameter combination. Optionally, when a variable is a meteorological parameter, the variable value can be the parameter value of the corresponding meteorological parameter; when a variable is a seasonal variable, the variable value can be the value of the seasonal variable.

[0062] Optionally, when determining the prediction performance index values ​​of the ozone exceedance prediction model under each combination of undetermined meteorological parameters, for any one of the multiple undetermined meteorological parameter combinations, the electronic device can acquire an ozone exceedance verification data set within the target area, and determine a parameter combination verification data set under any undetermined meteorological parameter combination based on the ozone exceedance verification data set. A parameter combination verification data set under any undetermined meteorological parameter combination may include an ozone exceedance verification label and the variable value of each independent variable under any undetermined meteorological parameter combination. The method for determining the parameter combination verification data set under any undetermined meteorological parameter combination can be the same as the method for determining the parameter combination training data set under the current undetermined meteorological parameter combination; this will not be elaborated further in this embodiment. Optionally, the electronic device's own storage space may store the ozone exceedance verification data set. In this case, the electronic device can acquire the ozone exceedance verification data set from its own storage space; or, the electronic device can acquire a verification data download link and use the verification data download link to download the ozone exceedance verification data set, etc.; this embodiment does not limit this. Accordingly, a set of ozone exceedance verification data may include, but is not limited to, an ozone exceedance verification label and a meteorological parameter data; for example, it may also include the time corresponding to the ozone exceedance verification data, or it may also include seasonal information indicating the season to which the time corresponding to the ozone exceedance verification data belongs, and so on.

[0063] Based on this, the electronic device can invoke the ozone exceedance prediction model under any given combination of meteorological parameters, and perform ozone exceedance prediction based on the verification data of each parameter combination in the verification dataset for any given combination of meteorological parameters, obtaining the ozone exceedance prediction probability value corresponding to each parameter combination verification data. Correspondingly, the electronic device can calculate the prediction performance index value of the ozone exceedance prediction model under any given combination of meteorological parameters based on the ozone exceedance prediction probability value corresponding to each parameter combination verification data and the ozone exceedance verification label. Here, the ozone exceedance probability indicated by a label can refer to the actual ozone exceedance probability within the corresponding time period.

[0064] In one implementation, when calculating the prediction performance index of the ozone exceedance prediction model under any unknown meteorological parameter combination based on the ozone exceedance prediction probability value and ozone exceedance verification label corresponding to each parameter combination verification data, the electronic device can determine the number of accurate prediction samples based on the ozone exceedance prediction probability value and ozone exceedance verification label corresponding to each parameter combination verification data. The accurate prediction sample can be the parameter combination verification data where the ozone exceedance situation indicated by the ozone exceedance prediction probability value is the same as the ozone exceedance situation indicated by the corresponding ozone exceedance verification label. That is, for any parameter combination verification data, when the ozone exceedance situation indicated by the ozone exceedance prediction probability value corresponding to any parameter combination verification data (e.g., if the ozone exceedance prediction probability value is greater than 0.5, the indicated ozone exceedance situation is ozone exceedance; if the ozone exceedance prediction probability value is less than or equal to 0.5, the indicated ozone exceedance situation is ozone not exceedance) is the same as the ozone exceedance situation indicated by the ozone exceedance verification label corresponding to any parameter combination verification data, any parameter combination verification data can be regarded as an accurate prediction sample. Therefore, the ratio between the number of accurately predicted samples and the number of parameter combination validation data in the parameter combination validation dataset for any given combination of meteorological parameters can be used as the prediction performance index of the ozone exceedance prediction model for any given combination of meteorological parameters. Here, the ozone exceedance validation label corresponding to a parameter combination validation dataset can be the ozone exceedance validation label within the corresponding parameter combination validation dataset. In this case, the larger the prediction performance index value, the better the prediction performance of the corresponding ozone exceedance prediction model.

[0065] In another implementation, the electronic device can combine the parameter combination verification data set under any unknown meteorological parameter combination based on the ozone exceedance verification tag corresponding to each parameter combination verification data set, to obtain multiple verification data pairs. Each verification data pair can include a positive sample and a negative sample. A positive sample can be a parameter combination verification data set indicating that the ozone exceedance situation indicated by the ozone exceedance verification tag is ozone exceedance, and a negative sample can be a parameter combination verification data set indicating that the ozone exceedance situation indicated by the ozone exceedance verification tag is ozone not exceedance. Then, based on the ozone exceedance prediction probability value corresponding to each parameter combination verification data set, the multiple verification data sets can be used to generate a prediction probability. Based on the identification of at least one target validation data pair, a target validation data pair can refer to a validation data pair in which the ozone exceedance prediction probability value of the positive sample is greater than that of the negative sample. That is, each validation data pair in which the ozone exceedance prediction probability value of the positive sample is greater than that of the negative sample can be regarded as a target validation data pair, so as to identify at least one target validation data pair. Accordingly, the ratio between the number of target validation data pairs in at least one target validation data pair and the number of validation data pairs in multiple validation data pairs can be used as the prediction performance index value of the ozone exceedance prediction model under any combination of undetermined meteorological parameters. In this context, the ratio between the number of target validation data pairs in at least one target validation data pair and the number of validation data pairs in multiple validation data pairs can be used as the AUC value of the ozone exceedance prediction model under any given combination of meteorological parameters. In other words, the electronic device can calculate the AUC value of the ozone exceedance prediction model under any given combination of meteorological parameters based on the ozone exceedance prediction probability value and ozone exceedance verification label corresponding to the validation data of each parameter combination, and use this value as the prediction performance index value of the ozone exceedance prediction model under any given combination of meteorological parameters. In this case, the larger the prediction performance index value, the better the prediction performance of the corresponding ozone exceedance prediction model.

[0066] In another implementation, the electronic device can also calculate the Brier-score (also known as the Brier score) of the ozone exceedance prediction model under any unknown meteorological parameter combination based on the ozone exceedance prediction probability value and ozone exceedance verification label corresponding to the verification data of each parameter combination, and use the Brier-score of the ozone exceedance prediction model under any unknown meteorological parameter combination as the prediction performance index value of the ozone exceedance prediction model under any unknown meteorological parameter combination, and so on. Optionally, the electronic device can use Formula 1.2 to calculate the Brier score of the ozone exceedance prediction model under any unknown meteorological parameter combination:

[0067] Equation 1.2

[0068] in, Let y represent the ozone exceedance prediction probability value of the t-th parameter combination validation data in the parameter combination validation data set under any unknown meteorological parameter combination. t Let represent the actual probability of ozone exceedance in the validation data of the t-th parameter combination (e.g., 1 for exceedance, 0 for non-exceedance), and T represent the number of parameter combination validation data in the set of parameter combination validation data under any unknown meteorological parameter combination. In this case, the smaller the prediction performance index value (i.e., the closer the predicted ozone exceedance value is to the corresponding actual probability of ozone exceedance), the better the prediction performance of the corresponding ozone exceedance prediction model.

[0069] S305. Based on the prediction performance index values ​​of the ozone exceedance prediction model under various combinations of undetermined meteorological parameters, the key meteorological parameter combinations are determined, and the ozone exceedance prediction model under the key meteorological parameter combinations is used as the target ozone exceedance prediction model for the target area. The target ozone exceedance prediction model is used to predict ozone exceedance according to the key meteorological parameter combinations.

[0070] Optionally, when determining key meteorological parameter combinations based on the prediction performance index values ​​of ozone exceedance prediction models under various combinations of undetermined meteorological parameters, the electronic device can determine the current target meteorological parameter combination from multiple combinations of undetermined meteorological parameters based on the prediction performance index values ​​of ozone exceedance prediction models under various combinations of undetermined meteorological parameters, and add the current target meteorological parameter combination to the target meteorological parameter combination set. Specifically, the combination of undetermined meteorological parameters corresponding to the ozone exceedance prediction model with the best prediction performance can be determined from multiple combinations of undetermined meteorological parameters, and the combination of undetermined meteorological parameters corresponding to the ozone exceedance prediction model with the best prediction performance can be used as the current target meteorological parameter combination. That is, the current target meteorological parameter combination can be the combination of undetermined meteorological parameters corresponding to the ozone exceedance prediction model with the best prediction performance among multiple combinations of undetermined meteorological parameters. Optionally, when a larger prediction performance index value indicates better prediction performance (i.e., superior performance), the current target meteorological parameter combination can be the undetermined meteorological parameter combination corresponding to the ozone exceedance prediction model with the largest prediction performance index value. Conversely, when a smaller prediction performance index value indicates better prediction performance, the current target meteorological parameter combination can be the undetermined meteorological parameter combination corresponding to the ozone exceedance prediction model with the smallest prediction performance index value. For example, such as... Figure 4 As shown, taking the prediction performance index AUC value as an example, assuming the initial number of meteorological parameters is 2, the AUC values ​​of the ozone exceedance prediction model under all combinations of undetermined meteorological parameters with 2 meteorological parameters can be obtained. In this case, when the AUC value of the ozone exceedance prediction model under the undetermined meteorological parameter combination consisting of temperature and vertical wind speed is the largest, the undetermined meteorological parameter combination consisting of temperature and vertical wind speed can be used as the current target meteorological parameter combination. In this embodiment of the invention, an ozone exceedance prediction model under a meteorological parameter combination corresponds to the corresponding meteorological parameter combination.

[0071] Then, the current target meteorological parameter combination and each meteorological parameter other than the current target meteorological parameter combination can be combined to obtain at least one updated meteorological parameter combination. For example, assuming that the current target meteorological parameter combination includes temperature and vertical wind speed, and the multiple meteorological parameters include temperature, specific humidity, cloud cover, total precipitation, pressure, horizontal wind speed, vertical wind speed, atmospheric boundary layer height, and ozone column concentration, then at least one updated meteorological parameter combination may include: [temperature, vertical wind speed, specific humidity], [temperature, vertical wind speed, cloud cover], [temperature, vertical wind speed, total precipitation], [temperature, vertical wind speed, pressure], [temperature, vertical wind speed, horizontal wind speed], [temperature, vertical wind speed, atmospheric boundary layer height], and [temperature, vertical wind speed, ozone column concentration].

[0072] Based on this, the electronic device can construct ozone exceedance prediction models for each of at least one updated meteorological parameter combination based on the ozone exceedance training data set, and determine the prediction performance index value of the ozone exceedance prediction model for each updated meteorological parameter combination. Then, based on the prediction performance index value of the ozone exceedance prediction model for each updated meteorological parameter combination, a target updated meteorological parameter combination is determined from at least one updated meteorological parameter combination, and the current target meteorological parameter combination is updated using the target updated meteorological parameter combination, thereby adding the current target meteorological parameter combination (which can be the updated current target meteorological parameter combination) to the target meteorological parameter combination set. The implementation method of determining the target updated meteorological parameter combination from at least one updated meteorological parameter combination is the same as the implementation method of determining the current target meteorological parameter combination from multiple undetermined meteorological parameter combinations, and the method of determining the prediction performance index value is the same as described above. The embodiments of the present invention will not be repeated here. Accordingly, the electronic device can iteratively execute the above-described combination of the current target meteorological parameter combination and each meteorological parameter other than the current target meteorological parameter combination from the multiple meteorological parameters, to obtain at least one updated meteorological parameter combination, until a stopping combination condition is met, so as to determine the key meteorological parameter combination from the set of target meteorological parameter combinations. Optionally, the stopping combination condition may refer to the set of target meteorological parameter combinations including a target meteorological parameter combination with the number of meteorological parameters equal to the number of meteorological parameters in the multiple meteorological parameters; or, it may refer to the set of target meteorological parameter combinations other than the current target meteorological parameter combination in the set of target meteorological parameter combinations where there is a target meteorological parameter combination whose ozone exceedance prediction model has a better predictive performance than the ozone exceedance prediction model under the current target meteorological parameter combination, etc.; the embodiments of the present invention do not limit this.

[0073] Optionally, when constructing ozone exceedance prediction models for each of the at least one updated meteorological parameter combinations based on the ozone exceedance training dataset, for any updated meteorological parameter combination in the at least one updated meteorological parameter combination, the electronic device can determine the parameter combination training dataset for any updated meteorological parameter combination based on the ozone exceedance training dataset; and determine the ozone exceedance prediction model to be fitted for any updated meteorological parameter combination based on the ozone exceedance prediction model under the current target meteorological parameter combination; thereby using the parameter combination training dataset under any updated meteorological parameter combination to fit the ozone exceedance prediction model to be fitted for any updated meteorological parameter combination, and obtaining the ozone exceedance prediction model under any updated meteorological parameter combination. Based on this, when determining the ozone exceedance prediction model to be fitted under any updated meteorological parameter combination based on the ozone exceedance prediction model under the current target meteorological parameter combination, the function (i.e., the fitted nonparametric function, or the determined function, etc.) of each meteorological parameter in the current target meteorological parameter combination can be determined from the ozone exceedance prediction model under the current target meteorological parameter combination. This determined function is then used as the function of the corresponding meteorological parameter in the ozone exceedance prediction model to be fitted under any updated meteorological parameter combination, thereby determining the ozone exceedance prediction model to be fitted under any updated meteorological parameter combination. In this case, the function of each meteorological parameter in the current target meteorological parameter combination in the ozone exceedance prediction model to be fitted under any updated meteorological parameter combination is a determined function (i.e., a fitted nonparametric function). Therefore, in this embodiment of the invention, it is only necessary to fit the nonparametric functions of the meteorological parameters other than those in the current target meteorological parameter combination, as well as other parameters to be fitted (such as intercepts), to achieve the fitting of the ozone exceedance prediction model to be fitted under any updated meteorological parameter combination. It should be noted that the specific implementation method for fitting the ozone exceedance prediction model under any updated meteorological parameter combination can be the same as the specific implementation method for fitting the ozone exceedance prediction model under the current unknown meteorological parameter combination, and the embodiments of the present invention will not be repeated here, etc.

[0074] Optionally, when determining the key meteorological parameter combination from the set of target meteorological parameter combinations, the electronic device may use the target meteorological parameter combination corresponding to the ozone exceedance prediction model with the best prediction performance in the set of target meteorological parameter combinations as the key meteorological parameter combination, thereby determining the key meteorological parameter combination from the set of target meteorological parameter combinations. For example, such as... Figure 5As shown, the AUC value is used as an example to illustrate the prediction performance index. A higher AUC value indicates better prediction performance. Assuming the target meteorological parameter combination set can include target meteorological parameter combinations with a number of meteorological parameters (i.e., the number of meteorological parameters) ranging from 1 to 9, and the AUC value is highest when the number of meteorological parameters is 8, then the target meteorological parameter combination with 8 meteorological parameters can be determined as the key meteorological parameter combination. Optionally, when the target meteorological parameter combination set includes multiple target meteorological parameter combinations with the highest AUC value, the key meteorological parameter combination can be the target meteorological parameter combination with the largest number of meteorological parameters among these multiple combinations with the highest AUC value, and so on; this embodiment of the invention does not limit this.

[0075] Based on this, the embodiments of the present invention can use the function of the already fitted meteorological parameters (i.e., the function of each meteorological parameter in the current target meteorological parameter combination in the ozone exceedance prediction model under the current target meteorological parameter combination) when constructing an ozone exceedance prediction model under any updated meteorological parameter combination. This avoids refitting the already fitted meteorological parameters, thereby achieving stepwise regression and effectively reducing the amount of computation. Furthermore, it does not require traversing all possible combinations of meteorological parameters from the initial meteorological quantity to the number of meteorological parameters among multiple meteorological parameters, thus eliminating the need to construct ozone exceedance prediction models under certain meteorological parameter combinations, thereby effectively reducing the amount of computation.

[0076] Optionally, in other embodiments, the multiple combinations of undetermined meteorological parameters may also include all possible combinations of meteorological parameters from the initial number of meteorological parameters to the number of meteorological parameters among the multiple meteorological parameters (such as 2-9, etc.). In this case, the key meteorological parameter combination can be directly determined from the multiple combinations of undetermined meteorological parameters, etc.; the present invention does not limit this.

[0077] Optionally, after obtaining the target ozone exceedance prediction model for the target area, the electronic device can also acquire the ozone exceedance test data set within the target area and call the target ozone exceedance prediction model respectively. Based on each ozone exceedance test data in the ozone exceedance test data set, ozone exceedance prediction is performed (i.e., ozone exceedance prediction is performed according to the combination of key meteorological parameters based on each ozone exceedance test data in the ozone exceedance test data set), obtaining the ozone exceedance prediction probability value corresponding to each ozone exceedance test data. Therefore, based on the ozone exceedance prediction probability value corresponding to each ozone exceedance test data and the ozone exceedance test label (which can be used to indicate the actual ozone exceedance probability), the test index value of the target ozone exceedance prediction model can be calculated. This test index value can be further used to indicate the prediction performance of the target ozone exceedance prediction model. Optionally, the test index value may include, but is not limited to, at least one of the following: Brier-score and AUC value, etc., which are not limited in this embodiment of the invention. It should be noted that the embodiments of the present invention do not limit the method of obtaining the ozone exceedance test dataset, nor the method of calculating the test index values ​​(such as the same method as the calculation method of the prediction performance index values). For example, taking one region as one city, calculations show that after removing cities with no data, the mean AUC value of the target ozone exceedance prediction model for all cities in the country is 0.89, and the mean Brier-score is 0.04. The prediction performance of the target ozone exceedance prediction model for each city is excellent, thus verifying the robustness of the target ozone exceedance prediction model.

[0078] Based on this, when a region is a city and a time is a day, the embodiments of the present invention can achieve ozone exceedance prediction at the urban spatial scale and daily time scale based on key meteorological parameters, thereby obtaining the corresponding ozone exceedance prediction results (such as ozone exceedance prediction probability values), so as to achieve accurate prediction of the probability of ozone exceedance under daily meteorological conditions at the urban spatial scale.

[0079] In summary, this invention can screen key meteorological parameters (i.e., meteorological parameters in the key meteorological parameter combination) for each region through stepwise regression and meteorological parameter combinations to avoid the possibility of model overfitting and underfitting. Based on the key meteorological parameters of each region, a target ozone exceedance prediction model for that region is determined, and ozone exceedance prediction is performed. The predictive performance of the model on the test set is verified based on the AUC value and / or Brier value (i.e., Brier score). This enables the construction of a meteorological-driven target ozone exceedance prediction model with excellent predictive performance at both the urban spatial scale and daily time scale. Correspondingly, this invention can analyze the relationship between meteorology and ozone exceedance, providing technical support and a scientific basis for management units to conduct assessments of good weather days and formulate relevant policies. It provides a scientific basis for relevant management units to prevent ozone exceedances and efficiently control ozone pollution, which is of great significance.

[0080] This invention, in its embodiments, after obtaining a training dataset of ozone exceedances within a target area, determines initial meteorological parameters. Multiple meteorological parameters are then combined according to these initial parameters to obtain multiple combinations of undetermined meteorological parameters, where the number of meteorological parameters in each undetermined combination equals the initial meteorological parameters. Based on the ozone exceedance training dataset, ozone exceedance prediction models are constructed for each of the multiple undetermined meteorological parameter combinations, and the prediction performance index values ​​of each undetermined meteorological parameter combination are determined. Furthermore, based on the prediction performance index values ​​of the ozone exceedance prediction models for each undetermined meteorological parameter combination, key meteorological parameter combinations are determined, and the ozone exceedance prediction models for these key meteorological parameter combinations are used as target ozone exceedance prediction models for the target area. These target ozone exceedance prediction models are then used to predict ozone exceedances according to the key meteorological parameter combinations. As can be seen, the embodiments of the present invention can screen out the key meteorological parameters affecting ozone exceedance in the target area, effectively avoid the possibility of overfitting and underfitting, optimize the predictive performance of the model, and thus obtain a target ozone exceedance prediction model with better predictive performance. This allows for convenient ozone exceedance prediction and can effectively improve the accuracy of ozone exceedance prediction.

[0081] Based on the description of the relevant embodiments of the ozone exceedance prediction method above, this invention also proposes an ozone exceedance prediction device, which can be a computer program (including program code) running in an electronic device; such as Figure 6 As shown, the ozone exceedance prediction device may include an acquisition unit 601 and a processing unit 602. The ozone exceedance prediction device can perform... Figure 1 or Figure 3 The ozone exceedance prediction method shown, i.e., the ozone exceedance prediction device can operate the above-mentioned unit:

[0082] The acquisition unit 601 is used to acquire a set of ozone exceedance training data in the target area. One ozone exceedance training data includes an ozone exceedance training label and a meteorological parameter data. One meteorological parameter data includes the parameter values ​​of each meteorological parameter among multiple meteorological parameters.

[0083] The processing unit 602 is used to construct ozone exceedance prediction models for each of the multiple combinations of undetermined meteorological parameters based on the ozone exceedance training data set; and to determine the prediction performance index values ​​of the ozone exceedance prediction models for each combination of undetermined meteorological parameters.

[0084] The processing unit 602 is further configured to determine key meteorological parameter combinations based on the prediction performance index values ​​of the ozone exceedance prediction models under each combination of undetermined meteorological parameters, and to use the ozone exceedance prediction models under the key meteorological parameter combinations as the target ozone exceedance prediction models for the target area. The target ozone exceedance prediction models are used to predict ozone exceedances according to the key meteorological parameter combinations.

[0085] In one embodiment, when the processing unit 602 constructs ozone exceedance prediction models for each of the multiple combinations of undetermined meteorological parameters based on the ozone exceedance training data set, it can specifically be used for:

[0086] Iterate through each of the multiple undetermined meteorological parameter combinations and take the currently iterated undetermined meteorological parameter combination as the current undetermined meteorological parameter combination.

[0087] Based on the ozone exceedance training data set, a parameter combination training data set under the current undetermined meteorological parameter combination is determined. The parameter combination training data under the current undetermined meteorological parameter combination includes an ozone exceedance training label and the variable value of each independent variable among the multiple independent variables under the current undetermined meteorological parameter combination. The multiple independent variables under the current undetermined meteorological parameter combination include each meteorological parameter and / or seasonal variable in the current undetermined meteorological parameter combination.

[0088] Using the parameter combination training data set under the current undetermined meteorological parameter combination, an ozone exceedance prediction model under the current undetermined meteorological parameter group is constructed;

[0089] After traversing through each of the multiple combinations of undetermined meteorological parameters, an ozone exceedance prediction model is obtained for each combination of undetermined meteorological parameters.

[0090] In another implementation, an ozone exceedance prediction model is a logical generalized additive model, and the ozone exceedance training label in an ozone exceedance training dataset is used to indicate whether the ozone exceeds the standard within the corresponding time period of the ozone exceedance training dataset; when the processing unit 602 constructs the ozone exceedance prediction model under the current undetermined meteorological parameter group using the parameter combination training dataset under the current undetermined meteorological parameter group, it can be specifically used for:

[0091] Determine the ozone exceedance prediction model to be fitted under the current unknown meteorological parameter combination, wherein the ozone exceedance prediction model to be fitted includes a nonparametric smoothing function of each meteorological parameter in the current unknown meteorological parameter combination;

[0092] Using the parameter combination training data set under the current undetermined meteorological parameter combination, the ozone exceedance prediction model to be fitted is fitted to obtain the ozone exceedance prediction model under the current undetermined meteorological parameter combination.

[0093] In another embodiment, when determining the prediction performance index values ​​of the ozone exceedance prediction model under each combination of undetermined meteorological parameters, the processing unit 602 may specifically be used for:

[0094] For any one of the multiple undetermined meteorological parameter combinations, obtain the ozone exceedance verification data set in the target area, and determine the parameter combination verification data set under any one undetermined meteorological parameter combination based on the ozone exceedance verification data set. The parameter combination verification data under any one undetermined meteorological parameter combination includes an ozone exceedance verification label and the variable value of each independent variable under any one undetermined meteorological parameter combination.

[0095] The ozone exceedance prediction model under any of the undetermined meteorological parameter combinations is called respectively. Based on the verification data of each parameter combination in the parameter combination verification data set under any of the undetermined meteorological parameter combinations, ozone exceedance prediction is performed to obtain the ozone exceedance prediction probability value corresponding to each parameter combination verification data.

[0096] Based on the ozone exceedance prediction probability value and ozone exceedance verification label corresponding to the verification data of each parameter combination, the prediction performance index value of the ozone exceedance prediction model under any unknown meteorological parameter combination is calculated.

[0097] In another embodiment, the processing unit 602 may also be used for:

[0098] Determine the initial meteorological quantities;

[0099] The multiple meteorological parameters are combined according to the initial meteorological quantity to obtain the multiple undetermined meteorological parameter combinations, wherein the number of meteorological parameters in an undetermined meteorological parameter combination is equal to the initial meteorological quantity.

[0100] In another embodiment, when determining the key meteorological parameter combinations based on the prediction performance index values ​​of the ozone exceedance prediction model under each combination of undetermined meteorological parameters, the processing unit 602 may specifically be used for:

[0101] Based on the prediction performance index values ​​of the ozone exceedance prediction model under each combination of undetermined meteorological parameters, the current target meteorological parameter combination is determined from the multiple combinations of undetermined meteorological parameters, and the current target meteorological parameter combination is added to the target meteorological parameter combination set.

[0102] The current target meteorological parameter combination and each meteorological parameter other than the current target meteorological parameter combination are combined respectively to obtain at least one updated meteorological parameter combination;

[0103] Based on the ozone exceedance training dataset, ozone exceedance prediction models are constructed for each of the at least one updated meteorological parameter combinations, and the prediction performance index values ​​of the ozone exceedance prediction models for each of the updated meteorological parameter combinations are determined.

[0104] Based on the prediction performance index values ​​of the ozone exceedance prediction model under each updated meteorological parameter combination, a target updated meteorological parameter combination is determined from the at least one updated meteorological parameter combination, and the current target meteorological parameter combination is updated using the target updated meteorological parameter combination, thereby adding the current target meteorological parameter combination to the target meteorological parameter combination set;

[0105] The process iteratively combines the current target meteorological parameter combination with each of the multiple meteorological parameters other than the current target meteorological parameter combination to obtain at least one updated meteorological parameter combination, until the combination stops, so as to determine the key meteorological parameter combination from the set of target meteorological parameter combinations.

[0106] In another embodiment, when the processing unit 602 constructs ozone exceedance prediction models for each of the at least one updated meteorological parameter combinations based on the ozone exceedance training data set, it can specifically be used to:

[0107] For any of the at least one updated meteorological parameter combinations, based on the ozone exceedance training data set, determine the parameter combination training data set under any of the updated meteorological parameter combinations;

[0108] Based on the ozone exceedance prediction model under the current target meteorological parameter combination, determine the ozone exceedance prediction model to be fitted under any updated meteorological parameter combination.

[0109] Using the parameter combination training data set under any of the updated meteorological parameter combinations, the ozone exceedance prediction model to be fitted under any of the updated meteorological parameter combinations is fitted to obtain the ozone exceedance prediction model under any of the updated meteorological parameter combinations.

[0110] According to one embodiment of the present invention, Figure 6Each unit in the ozone exceedance prediction device shown can be individually or entirely combined into one or more other units, or one or more of the units can be further divided into multiple functionally smaller units. This achieves the same operation without affecting the technical effect of the embodiments of the present invention. The above units are based on logical function division. In practical applications, the function of one unit can be implemented by multiple units, or the function of multiple units can be implemented by one unit. In other embodiments of the present invention, any ozone exceedance prediction device may also include other units. In practical applications, these functions can also be implemented with the assistance of other units, and can be implemented collaboratively by multiple units.

[0111] According to another embodiment of the present invention, it is possible to perform operations such as those described above by running on a general-purpose electronic device, such as a computer, which includes processing elements and storage elements such as a central processing unit (CPU), random access memory (RAM), and read-only memory (ROM). Figure 1 or Figure 3 The computer program (including program code) involved in each step of the corresponding method shown, to construct such... Figure 6 The ozone exceedance prediction device shown herein, and the ozone exceedance prediction method for implementing embodiments of the present invention, are described. The computer program may be recorded on, for example, a computer storage medium, loaded onto the aforementioned electronic device via the computer storage medium, and run therein.

[0112] This invention provides an embodiment of ozone exceedance training data set within a target area. Each ozone exceedance training data set includes an ozone exceedance training label and meteorological parameter data, with each meteorological parameter data set including the parameter values ​​of various meteorological parameters. Then, based on the ozone exceedance training data set, ozone exceedance prediction models can be constructed for each of the multiple combinations of undetermined meteorological parameters; and the prediction performance index value of each ozone exceedance prediction model under each combination of undetermined meteorological parameters can be determined. Furthermore, based on the prediction performance index values ​​of the ozone exceedance prediction models under each combination of undetermined meteorological parameters, key meteorological parameter combinations can be determined, and the ozone exceedance prediction model under the key meteorological parameter combinations can be used as the target ozone exceedance prediction model for the target area. The target ozone exceedance prediction model is used to predict ozone exceedances according to the key meteorological parameter combinations. As can be seen, the embodiments of the present invention can conveniently predict ozone exceedances, effectively reduce the computational cost and time consumption of ozone exceedance prediction; furthermore, the embodiments of the present invention can determine the combination of key meteorological parameters to determine the target ozone exceedance prediction model for the target area, thereby enabling ozone exceedance prediction to be performed according to the combination of key meteorological parameters using the target ozone exceedance prediction model, which can effectively improve the accuracy of ozone exceedance prediction, thus improving the accuracy of ozone exceedance prediction for the target area.

[0113] Based on the description of the method and apparatus embodiments above, an exemplary embodiment of the present invention also provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores a computer program executable by the at least one processor, which, when executed by the at least one processor, causes the electronic device to perform the method according to an embodiment of the present invention.

[0114] An exemplary embodiment of the present invention also provides a non-transitory computer-readable storage medium storing a computer program, wherein the computer program, when executed by a computer's processor, is used to cause the computer to perform a method according to an embodiment of the present invention.

[0115] An exemplary embodiment of the present invention also provides a computer program product, including a computer program, wherein, when executed by a computer's processor, the computer program is used to cause the computer to perform a method according to an embodiment of the present invention.

[0116] refer to Figure 7 The present invention will now be described in the form of a structural block diagram of an electronic device 700 that can serve as a server or client of the present invention, which is an example of a hardware device that can be applied to various aspects of the present invention. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0117] like Figure 7 As shown, the electronic device 700 includes a computing unit 701, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 702 or a computer program loaded from a storage unit 708 into a random access memory (RAM) 703. The RAM 703 may also store various programs and data required for the operation of the electronic device 700. The computing unit 701, ROM 702, and RAM 703 are interconnected via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0118] Multiple components in electronic device 700 are connected to I / O interface 705, including: input unit 706, output unit 707, storage unit 708, and communication unit 709. Input unit 706 can be any type of device capable of inputting information to electronic device 700. Input unit 706 can receive input digital or character information and generate key signal inputs related to user settings and / or function control of electronic device. Output unit 707 can be any type of device capable of presenting information and may include, but is not limited to, a display, speaker, video / audio output terminal, vibrator, and / or printer. Storage unit 708 may include, but is not limited to, disk and optical disk. Communication unit 709 allows electronic device 700 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks, and may include, but is not limited to, modems, network cards, infrared communication devices, wireless communication transceivers, and / or chipsets, such as Bluetooth™ devices, WiFi devices, WiMax devices, cellular communication devices, and / or the like.

[0119] The computing unit 701 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 701 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 701 performs the various methods and processes described above. For example, in some embodiments, the ozone exceedance prediction method can be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 708. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 700 via ROM 702 and / or communication unit 709. In some embodiments, the computing unit 701 can be configured to perform the ozone exceedance prediction method by any other suitable means (e.g., by means of firmware).

[0120] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0121] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0122] As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or apparatus (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including machine-readable media that receive machine instructions as machine-readable signals. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0123] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0124] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0125] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other.

[0126] Furthermore, it should be understood that the above-disclosed embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, any equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. An ozone over standard prediction method, characterized by, The method comprises the following steps: obtaining an ozone over-limit training data set in a target area, one ozone over-limit training data comprising one ozone over-limit training label and one meteorological parameter data, one meteorological parameter data comprising parameter values of each meteorological parameter in multiple meteorological parameters; based on the ozone over-limit training data set, ozone over-limit prediction models under each of multiple to-be-determined meteorological parameter combinations are respectively constructed; and prediction performance index values of the ozone over-limit prediction models under the each of the to-be-determined meteorological parameter combinations are respectively determined, wherein one ozone over-limit prediction model is a logistic generalized additive model, and the ozone over-limit prediction model comprises a non-parametric smoothing function of each meteorological parameter in the corresponding to-be-determined meteorological parameter combination, which is determined by separately modeling each meteorological parameter and ozone over-limit probability under the corresponding to-be-determined meteorological parameter combination; based on the prediction performance index values of the ozone over-limit prediction models under the each of the to-be-determined meteorological parameter combinations, a key meteorological parameter combination is determined, and the ozone over-limit prediction model under the key meteorological parameter combination is taken as a target ozone over-limit prediction model of the target area, which is used for ozone over-limit prediction according to the key meteorological parameter combination; The method further comprises: determining an initial meteorological quantity; combining the multiple meteorological parameters according to the initial meteorological quantity to obtain the multiple to-be-determined meteorological parameter combinations, and the number of meteorological parameters in one to-be-determined meteorological parameter combination is equal to the initial meteorological quantity; wherein, based on the prediction performance index values of the ozone over-limit prediction models under the each of the to-be-determined meteorological parameter combinations, the key meteorological parameter combination is determined, which comprises: based on the prediction performance index values of the ozone over-limit prediction models under the each of the to-be-determined meteorological parameter combinations, a current target meteorological parameter combination is determined from the multiple to-be-determined meteorological parameter combinations, and the current target meteorological parameter combination is added to a target meteorological parameter combination set; each of the current target meteorological parameter combination and each meteorological parameter except the current target meteorological parameter combination in the multiple meteorological parameters is combined to obtain at least one updated meteorological parameter combination; based on the ozone over-limit training data set, ozone over-limit prediction models under each of the at least one updated meteorological parameter combination are respectively constructed, and prediction performance index values of the ozone over-limit prediction models under the each of the at least one updated meteorological parameter combination are respectively determined; based on the prediction performance index values of the ozone over-limit prediction models under the each of the at least one updated meteorological parameter combination, a target updated meteorological parameter combination is determined from the at least one updated meteorological parameter combination to update the current target meteorological parameter combination with the target updated meteorological parameter combination, so that the current target meteorological parameter combination is added to the target meteorological parameter combination set; iteratively performing the combining of the current target meteorological parameter combination and each meteorological parameter in the plurality of meteorological parameters except the current target meteorological parameter combination respectively to obtain at least one updated meteorological parameter combination until a stop combining condition is reached to determine a key meteorological parameter combination from the set of target meteorological parameter combinations; the ozone over-limit training data set, respectively constructing an ozone over-limit prediction model under each of a plurality of to-be-determined meteorological parameter combinations, comprising: traversing each of the plurality of to-be-determined meteorological parameter combinations, and taking the currently traversed to-be-determined meteorological parameter combination as a current to-be-determined meteorological parameter combination; based on the ozone over-limit training data set, determining a parameter combination training data set under the current to-be-determined meteorological parameter combination, one parameter combination training data under the current to-be-determined meteorological parameter combination comprising one ozone over-limit training label and a variable value of each independent variable under the current to-be-determined meteorological parameter combination, the plurality of independent variables under the current to-be-determined meteorological parameter combination comprising each meteorological parameter in the current to-be-determined meteorological parameter combination and / or a seasonal variable; using the parameter combination training data set under the current to-be-determined meteorological parameter combination, constructing an ozone over-limit prediction model under the current to-be-determined meteorological parameter combination; after traversing each of the plurality of to-be-determined meteorological parameter combinations, obtaining the ozone over-limit prediction model under each of the plurality of to-be-determined meteorological parameter combinations; an ozone over-limit training label in one ozone over-limit training data is used to indicate whether the ozone is over-limit in the time corresponding to the corresponding ozone over-limit training data; the using the parameter combination training data set under the current to-be-determined meteorological parameter combination, constructing an ozone over-limit prediction model under the current to-be-determined meteorological parameter combination, comprising: determining a to-be-fitted ozone over-limit prediction model under the current to-be-determined meteorological parameter combination; using the parameter combination training data set under the current to-be-determined meteorological parameter combination, fitting the to-be-fitted ozone over-limit prediction model to obtain the ozone over-limit prediction model under the current to-be-determined meteorological parameter combination; the respectively determining the prediction performance index value of the ozone over-limit prediction model under each of the plurality of to-be-determined meteorological parameter combinations, comprising: for any to-be-determined meteorological parameter combination in the plurality of to-be-determined meteorological parameter combinations, obtaining an ozone over-limit verification data set in the target region, and based on the ozone over-limit verification data set, determining a parameter combination verification data set under the any to-be-determined meteorological parameter combination, one parameter combination verification data under the any to-be-determined meteorological parameter combination comprising one ozone over-limit verification label and a variable value of each independent variable under the any to-be-determined meteorological parameter combination; respectively calling the ozone over-limit prediction model under the any to-be-determined meteorological parameter combination, based on each parameter combination verification data in the parameter combination verification data set under the any to-be-determined meteorological parameter combination to perform ozone over-limit prediction, obtaining an ozone over-limit prediction probability value corresponding to each parameter combination verification data; verify the ozone over-limit prediction probability value and the ozone over-limit verification label corresponding to each parameter combination verification data, and calculate a prediction performance index value of the ozone over-limit prediction model under any to-be-determined meteorological parameter combination; The method further includes: For any update meteorological parameter combination in the at least one update meteorological parameter combination, determining a parameter combination training data set under the any update meteorological parameter combination based on the ozone over-limit training data set; Based on the ozone over-limit prediction model under the current target meteorological parameter combination, determining a to-be-fitted ozone over-limit prediction model under the any update meteorological parameter combination; Fitting the to-be-fitted ozone over-limit prediction model under the any update meteorological parameter combination by using the parameter combination training data set under the any update meteorological parameter combination, to obtain the ozone over-limit prediction model under the any update meteorological parameter combination.

2. The ozone over-limit prediction device applied to the ozone over-limit prediction method of claim 1, characterized in that, The apparatus includes: An obtaining unit, configured to obtain an ozone over-limit training data set in a target region, one ozone over-limit training data including one ozone over-limit training label and one meteorological parameter data, one meteorological parameter data including parameter values of each meteorological parameter in a plurality of meteorological parameters; A processing unit, configured to construct an ozone over-limit prediction model under each to-be-determined meteorological parameter combination in a plurality of to-be-determined meteorological parameter combinations based on the ozone over-limit training data set respectively, and determine a prediction performance index value of the ozone over-limit prediction model under each to-be-determined meteorological parameter combination respectively, wherein one ozone over-limit prediction model is a logistic generalized additive model, and the ozone over-limit prediction model includes a non-parametric smooth function of each meteorological parameter in a corresponding to-be-determined meteorological parameter combination, the non-parametric smooth function being determined by separately modeling each meteorological parameter and an ozone over-limit probability under the corresponding to-be-determined meteorological parameter combination; The processing unit is further configured to determine a key meteorological parameter combination based on the prediction performance index values of the ozone over-limit prediction models under the to-be-determined meteorological parameter combinations, and take the ozone over-limit prediction model under the key meteorological parameter combination as a target ozone over-limit prediction model of the target region, the target ozone over-limit prediction model being used for ozone over-limit prediction according to the key meteorological parameter combination; The processing unit is further configured to determine an initial meteorological quantity, and combine the plurality of meteorological parameters according to the initial meteorological quantity to obtain the plurality of to-be-determined meteorological parameter combinations, a number of meteorological parameters in one to-be-determined meteorological parameter combination being equal to the initial meteorological quantity; When determining the key meteorological parameter combination based on the prediction performance index values of the ozone over-limit prediction models under the to-be-determined meteorological parameter combinations, the processing unit is further configured to: determine a current target meteorological parameter combination from the plurality of undetermined meteorological parameter combinations based on the prediction performance index values of the ozone exceedance prediction models under the respective undetermined meteorological parameter combinations, and add the current target meteorological parameter combination to a target meteorological parameter combination set; combine the current target meteorological parameter combination and each meteorological parameter except the current target meteorological parameter combination in the plurality of meteorological parameters respectively to obtain at least one updated meteorological parameter combination; construct an ozone exceedance prediction model under each updated meteorological parameter combination in the at least one updated meteorological parameter combination based on the ozone exceedance training data set, and determine a prediction performance index value of the ozone exceedance prediction model under each updated meteorological parameter combination respectively; determine a target updated meteorological parameter combination from the at least one updated meteorological parameter combination based on the prediction performance index values of the ozone exceedance prediction models under the respective updated meteorological parameter combinations, and update the current target meteorological parameter combination by using the target updated meteorological parameter combination, so as to add the current target meteorological parameter combination to the target meteorological parameter combination set; iteratively perform the combining the current target meteorological parameter combination and each meteorological parameter except the current target meteorological parameter combination in the plurality of meteorological parameters respectively to obtain at least one updated meteorological parameter combination until a stop combination condition is reached, so as to determine a key meteorological parameter combination from the target meteorological parameter combination set.

3. An electronic device, comprising: comprise: a processor; and a memory storing programs, wherein the programs include instructions that, when executed by the processor, cause the processor to perform the method according to claim 1.

4. A non-transitory computer readable storage medium having stored thereon computer instructions, wherein, The computer instructions are used to cause a computer to perform the method according to claim 1. The computer instructions are used to cause a computer to perform the method according to claim 1.

Images