Prediction and management method and system for multi-dimensional pollen information

Abstract

The present invention provides a prediction and management method and system suitable for multi-dimensional information of pollen. The method comprises the following steps: statistically analyzing historical project data of a display area and inputting them into a plurality of first-level models in an initial prediction model for training to obtain intermediate prediction results, combining and inputting the intermediate prediction results into a second-level model for training to generate a pollen prediction model; receiving a plurality of allergy factors set by a user terminal, obtaining real-time data of a display area based on the allergy factors and inputting the data into a pollen prediction model to obtain pollen concentration data, and inputting the pollen concentration data into a pollen health impact model to calculate the pollen impact logarithm; when the pollen impact logarithm is greater than a preset impact value, generating initial diffusion areas corresponding to respective planting areas based on an allergen display strategy and the planting areas of respective allergy factors in the display area; receiving regional data of the display area, adjusting the initial diffusion area to determine the allergen area, and generating early warning information based on the real-time location information of the user terminal and the allergen area and sending the information to the user terminal.

Description

Technical Field

[0001] The present invention relates to data processing technology, and in particular to a prediction and management method and system suitable for multi-dimensional pollen information. Background Art

[0002] With the continuous acceleration of urbanization and changes in the ecological environment, pollen allergy has become a major problem affecting public health. Every year during the pollen season, the pollen concentration in many areas rises sharply, causing a large number of allergy patients to experience symptoms. In order to effectively respond to this challenge, it is particularly important to accurately predict pollen concentrations and evaluate their impact on public health. The development of modern science and technology has provided us with a variety of data sources, including historical meteorological data, geographic information and real-time environmental data. The combination of these data can provide important support for the prediction and management of pollen concentrations.

[0003] At present, pollen concentration prediction mainly relies on simple statistical models and a single data source, with low prediction accuracy. Traditional methods usually cannot make full use of multi-dimensional historical data and real-time data, resulting in insufficient reliability and accuracy of the prediction results. For example, existing methods may only consider meteorological data, but ignore geographic information and the laws of historical pollen concentration changes. These problems seriously affect the actual application effect of pollen concentration prediction.

[0004] How to accurately display and forecast pollen-affected areas based on the actual situation in the region and provide more effective protection for public health has become an urgent problem that needs to be solved. Summary of the invention

[0005] The embodiments of the present invention provide a prediction and management method and system suitable for multi-dimensional pollen information, which can accurately display and forecast the pollen-affected area according to the actual situation of the area, thereby providing more effective protection for public health.

[0006] A first aspect of an embodiment of the present invention provides a method for predicting and managing multi-dimensional pollen information, comprising:

[0007] The historical project data of the statistical display area is used as an initial training set, and the initial training set is input into multiple primary models in the initial forecast model for training to obtain intermediate prediction results and combine them, and the intermediate training set is input into the secondary model for training to generate a pollen forecast model;

[0008] Receive multiple allergy factors set by the user end, obtain real-time data of the corresponding display area based on the allergy factors, input the real-time data into a pollen prediction model to obtain pollen concentration data, input the pollen concentration data into a pollen health impact model for calculation, and obtain the pollen impact logarithm;

[0009] When it is determined that the pollen impact logarithm is greater than the preset impact value, based on the allergen display strategy and the planting areas of each allergen in the display area, an initial diffusion area corresponding to each planting area in the display area is generated;

[0010] Receive regional data of the display area, customize and adjust the initial diffusion area according to the regional data to obtain an allergen area corresponding to the display area, and generate warning information based on the real-time location information of the user terminal and the allergen area and send it to the user terminal.

[0011] Optionally, in a possible implementation of the first aspect, the historical project data of the statistical display area is used as an initial training set, the initial training set is input into multiple primary models in the initial forecast model for training to obtain intermediate prediction results and the intermediate training sets are combined to obtain intermediate training sets and input into the secondary model for training to generate a pollen forecast model, including:

[0012] Counting historical project data of the display area as an initial training set, wherein the historical project data at least includes historical basic data and historical meteorological data;

[0013] Inputting the initial training set into multiple first-level models in the initial forecast model for training, and obtaining intermediate prediction results corresponding to each of the first-level models;

[0014] Combining the intermediate prediction results to obtain an intermediate training set, which is input into the secondary model for training, to obtain model weights corresponding to each of the primary models;

[0015] The model weights are configured to the corresponding primary models to generate a pollen prediction model.

[0016] Optionally, in a possible implementation of the first aspect, the receiving of multiple allergy factors set by a user terminal, acquiring real-time data of the corresponding display area based on the allergy factors, inputting the real-time data into a pollen prediction model to obtain pollen concentration data, and inputting the pollen concentration data into a pollen health impact model for calculation to obtain a pollen impact logarithm, including:

[0017] Receiving multiple allergy factors set by a user terminal, and acquiring real-time data of the corresponding display area based on the allergy factors;

[0018] Input the real-time data into the pollen prediction model to obtain the real-time prediction results corresponding to each of the first-level models, and obtain the pollen concentration data corresponding to the display area according to the real-time prediction results and the corresponding model weights;

[0019] The pollen concentration data is input into the pollen health impact model for calculation to obtain the pollen impact logarithm.

[0020] Optionally, in a possible implementation of the first aspect, inputting the pollen concentration data into a pollen health impact model for calculation to obtain a pollen impact logarithm includes:

[0021] The pollen impact logarithm is obtained by the following formula:

[0022] ;

[0023] in, is the logarithm of pollen impact, For the The number of outpatient visits per day, For the The expected number of outpatient visits per day, is the intercept, is the period dummy variable, is a nonparametric smoothing spline function, is a nonparametric plate spline function, For the date, is the pollen concentration, is the collaborative meteorological element variable, is a non-coordinated meteorological element variable, is the pollutant concentration, is the degree of freedom;

[0024] The pollen impact logarithm is sent to a management terminal.

[0025] Optionally, in a possible implementation of the first aspect, when it is determined that the pollen influence logarithm is greater than a preset influence value, based on the allergen display strategy and the planting areas of each allergen factor in the display area, generating an initial diffusion area corresponding to each planting area in the display area includes:

[0026] When it is determined that the pollen influence logarithm is greater than the preset influence value, the planting areas of each allergy factor in the display area and the sub-planting areas corresponding to each planting area are obtained;

[0027] The total planting area corresponding to the display area is obtained, and the planting proportion corresponding to each of the allergic factors is obtained according to the ratio of the sub-planting area to the total planting area;

[0028] Based on the product of the planting proportion and the corresponding pollen concentration data, sub-pollen concentration data corresponding to each of the planting areas is obtained;

[0029] According to the ratio of the sub-pollen concentration data and the preset area concentration data, a concentration diffusion coefficient is obtained, and based on the product of the concentration diffusion coefficient and the preset diffusion distance, a theoretical diffusion distance corresponding to the planting area is obtained;

[0030] The pixel points of the area outline of the planting area are moved based on the theoretical diffusion distance to generate initial diffusion areas corresponding to each planting area in the display area.

[0031] Optionally, in a possible implementation manner of the first aspect, the moving processing of the area outline pixel points of the planting area based on the theoretical diffusion distance to generate the initial diffusion area corresponding to each planting area in the display area includes:

[0032] Selecting any pixel point in the display area as a coordinate origin to construct a display coordinate system, and performing coordinate processing on the display area based on the display coordinate system;

[0033] Acquire the outline pixel points of each planting area in the display area as the area outline pixel points, and determine the area center coordinates of the planting area and the area center point corresponding to the area center coordinates according to the maximum abscissa, minimum abscissa, maximum ordinate and minimum ordinate of the area outline pixel points;

[0034] Sequentially connecting the center point of the region with the region outline pixel points of the corresponding planting region to obtain a diffusion connection line;

[0035] Based on the theoretical diffusion distance, the region outline pixel points are moved along the diffusion connection line and in a direction away from the region center point to obtain diffused region outline pixel points;

[0036] Adjacent pixel points of the diffused area contours are connected to obtain initial diffusion areas corresponding to each planting area.

[0037] Optionally, in a possible implementation of the first aspect, the method further includes:

[0038] Determine a concentration pixel value corresponding to the sub-pollen concentration data according to the sub-pollen concentration data and a preset concentration comparison table, wherein the preset concentration comparison table has a one-to-one correspondence between preset concentration intervals and concentration pixel values;

[0039] Filling the concentration pixel value into the corresponding planting area, and obtaining a gradient area according to the difference between the initial diffusion area and the planting area;

[0040] Sequentially moving the region contour pixel points along the diffusion connection line and in a direction away from the region center point based on a preset gradient distance to obtain gradient contour pixel points;

[0041] Connecting adjacent gradient contour pixel points to obtain gradient contour lines in a gradient region, segmenting the gradient region according to the gradient contour lines to obtain a plurality of gradient sub-regions, and sorting the gradient sub-regions along the diffusion connection line and in a direction away from the center point of the region to obtain a gradient sub-sequence;

[0042] Based on the ratio of the theoretical diffusion distance to the gradient distance, the gradient times are obtained;

[0043] The concentration pixel value is sequentially increased according to the preset attenuation value according to the number of gradients to obtain a plurality of gradient pixel values, and the gradient pixel values ​​are sequentially filled into the gradient sub-regions in the gradient sub-sequence.

[0044] Optionally, in a possible implementation of the first aspect, the method further includes:

[0045] When it is determined that the initial diffusion area has an overlapping area, the corresponding overlapping area is used as a fused diffusion area, and the corresponding initial diffusion area is used as an impact diffusion area, and the sub-pollen concentration data of the impact diffusion area is obtained as the impact concentration data;

[0046] According to the difference set between the affected diffusion area and the corresponding overlapping area, an independent diffusion area is obtained, and the affected concentration data corresponding to the affected diffusion area is used as the independent pollen concentration data of the independent diffusion area;

[0047] Acquire a plurality of regional pixel values ​​corresponding to the fused diffusion region, and the regional area corresponding to each of the regional pixel values;

[0048] Calculating the ratio of the area of ​​each region corresponding to the pixel value of each region to obtain the area ratio of the pixel value of the region;

[0049] A calculation is performed based on the area proportion and the corresponding pixel value of the region to obtain a fused pixel value corresponding to the fused diffusion region.

[0050] Optionally, in a possible implementation manner of the first aspect, the receiving the area data of the display area, customizing and adjusting the initial diffusion area according to the area data to obtain an allergen area corresponding to the display area, and generating warning information based on the real-time location information of the user terminal and the allergen area and sending it to the user terminal includes:

[0051] receiving regional data of the display area, the regional data including temperature, rainfall and wind speed;

[0052] The theoretical diffusion distance is adjusted according to the temperature, the rainfall and the wind speed to obtain an actual diffusion distance;

[0053] Based on the actual diffusion distance, pixel points of the area contour of the planting area are moved to generate an allergen area corresponding to the display area;

[0054] The real-time location information of the user terminal is obtained, and it is determined that the real-time location information is in the allergen area, and a warning message is generated and sent to the user terminal.

[0055] A second aspect of an embodiment of the present invention provides a prediction and management system for pollen multi-dimensional information, including:

[0056] A training module is used to collect historical project data of the display area as an initial training set, input the initial training set into multiple first-level models in the initial forecast model for training to obtain intermediate prediction results and combine them, obtain the intermediate training set and input it into the second-level model for training to generate a pollen forecast model;

[0057] A calculation module, for receiving a plurality of allergy factors set by a user terminal, obtaining real-time data of the corresponding display area based on the allergy factors, inputting the real-time data into a pollen prediction model to obtain pollen concentration data, and inputting the pollen concentration data into a pollen health impact model for calculation to obtain a pollen impact logarithm;

[0058] A generating module, for generating an initial diffusion area corresponding to each planting area in the display area based on the allergen display strategy and the planting area of ​​each allergen in the display area when it is determined that the pollen influence logarithm is greater than a preset influence value;

[0059] The adjustment module is used to receive the regional data of the display area, customize the initial diffusion area according to the regional data, obtain the allergen area corresponding to the display area, and generate warning information based on the real-time location information of the user terminal and the allergen area and send it to the user terminal.

[0060] According to a third aspect of an embodiment of the present invention, there is provided an electronic device, comprising: a memory, a processor and a computer program, wherein the computer program is stored in the memory, and the processor runs the computer program to execute the first aspect of the present invention and various methods that may be involved in the first aspect.

[0061] According to a fourth aspect of an embodiment of the present invention, a storage medium is provided, in which a computer program is stored. When the computer program is executed by a processor, it is used to implement the first aspect of the present invention and various methods that may be involved in the first aspect.

[0062] The beneficial effects of the present invention are as follows:

[0063] 1. The present invention can accurately display and forecast the pollen-affected area according to the actual situation of the area, providing more effective protection for public health. First of all, the present invention can more accurately predict the real-time pollen concentration through the hierarchical training of the initial forecast model and the secondary model. In addition, the present invention can conduct targeted pollen concentration information and health impact analysis based on the user's own allergic factors, conduct personalized health impact assessment for the user, and enhance the user experience. Secondly, the present invention can monitor environmental changes in real time, and notify the user in time when the pollen concentration exceeds the preset value, thereby reducing the impact of pollen allergies on the user's health and effectively reducing the risk of pollen allergies. Finally, the present invention can ensure the accuracy of the early warning information through customized adjustment of the initial diffusion area, so that users can better understand the distribution of allergens around them. Through the above measures, the present invention realizes the comprehensive management and precise control of pollen information, and provides users with effective protection means.

[0064] 2. The present invention can make intelligent predictions of pollen concentration and health impact, improve the accuracy and reliability of pollen impact forecast results, and provide users with timely and personalized health risk warnings. Among them, the present invention can obtain real-time data of the corresponding area, and then input it into the corresponding prediction model, improve the accuracy of pollen concentration through multi-level model training and optimization, and further input it into the pollen health impact model for calculation, obtain the pollen impact logarithm, and send it to the management end for forecasting in time, so as to provide personnel with allergy risk reminders and reduce allergy risks.

[0065] 3. The present invention can accurately identify high-risk areas by comprehensively analyzing the environmental data of the display area and the real-time location of the user, and provide timely warning information, so that users can take effective protective measures to reduce the risk of allergies, thereby improving the user experience and safety. Among them, the present invention can customize the pollen diffusion area corresponding to the planting area according to the actual data of the area, thereby improving the accuracy of the allergen area, and further identify the real-time location information of the user end, so as to send warning information, provide personnel with timely prevention of risk enhancement, and enable personnel to take protective measures to reduce the risk of allergies. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 A flowchart of a method for predicting and managing multi-dimensional pollen information provided by the present invention;

[0067] Figure 2 A schematic diagram of a gradual sub-region provided by the present invention;

[0068] Figure 3 A schematic diagram of a fusion diffusion region provided by the present invention;

[0069] Figure 4 A schematic diagram of the structure of a prediction and management system for pollen multi-dimensional information provided by the present invention;

[0070] Figure 5 A schematic diagram of the hardware structure of an electronic device provided by the present invention. DETAILED DESCRIPTION

[0071] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0072] like Figure 1 As shown, the present invention provides a flowchart of a method for predicting and managing multi-dimensional pollen information, which includes:

[0073] S1, statistically displaying historical project data of the area as an initial training set, inputting the initial training set into multiple first-level models in the initial forecast model for training to obtain intermediate prediction results and combining them, obtaining the intermediate training set and inputting it into the second-level model for training to generate a pollen forecast model.

[0074] It should be noted that in order to obtain the historical patterns of changes in pollen concentration in the area, statistical historical data can provide basic training data for the prediction model. The first-level model in the initial prediction model is responsible for preliminary analysis and prediction of data of different dimensions. The accuracy of the prediction is improved by combining multiple first-level models. The intermediate prediction results provide a preliminary prediction of pollen concentration. By inputting the intermediate training set into the second-level model and outputting a weight value, the prediction results are further optimized and corrected to generate a more accurate pollen prediction model.

[0075] It can be understood that the display area is the area where pollen density distribution is displayed, such as the area after the administrative area A city is divided according to the streets, namely, District A1, District A2 and District A3, etc. The historical project set is a collection of historical data, such as a collection of multi-dimensional data such as geographic data and meteorological data. The initial training set is a data set used for the initial training model. The initial forecast model is a model for initially predicting pollen information, including multiple first-level models, among which the first-level model can be an existing random forest algorithm, XGBoost, LightGBM, CatBoost, MLP-deep neural network, integration method and other algorithm models. The intermediate prediction result is the intermediate result output by the first-level model, which is used as the data result of input to continue training the second-level model. The second-level model is the advanced part of the model, which is responsible for optimizing the primary analysis results. The corresponding second-level model may be the same as the first-level model or different. The pollen forecast model is the final generated model, which is a Stacking model for predicting pollen concentration.

[0076] Through the above-mentioned implementation manner, the present invention can make full use of historical data through a multi-level training method to improve the accuracy and reliability of the divided concentration prediction.

[0077] In some embodiments, in step S1 (the historical project data of the statistical display area is used as an initial training set, the initial training set is input into multiple primary models in the initial forecast model for training to obtain intermediate prediction results and combined, the intermediate training set is input into the secondary model for training, and the pollen forecast model is generated), the specific implementation method includes:

[0078] S11, collecting historical project data of the statistical display area as an initial training set, wherein the historical project data at least includes historical basic data and historical meteorological data.

[0079] It is understandable that historical project data are statistically analyzed to obtain an initial training set so that the initial forecast model can be subsequently trained with data to improve the accuracy of the model forecast.

[0080] Among them, historical project data include historical basic data and historical meteorological data. Historical basic data refers to the basic data of the corresponding display area within the historical time period, such as basic geographic information, NDVI information output by satellite remote sensing, and other information. Historical meteorological data refers to meteorological data within the historical time period, such as humidity, temperature, wind speed and other meteorological data.

[0081] Through the above implementation, the present invention can obtain an initial training set to provide accurate training data for the initial prediction model.

[0082] S12, inputting the initial training set into multiple first-level models in the initial forecast model for training, and obtaining intermediate prediction results corresponding to each of the first-level models.

[0083] It can be understood that the initial training set is input into multiple first-level models in the initial prediction model in order to obtain intermediate prediction results predicted by multiple models, so as to facilitate subsequent combined input into the second-level model for optimization training in order to obtain accurate model weights, thereby improving the accuracy of pollen forecast results.

[0084] S13, combining the intermediate prediction results to obtain an intermediate training set which is input into the secondary model for training to obtain model weights corresponding to each of the primary models.

[0085] It can be understood that the intermediate training set is a data set of intermediate training results, which is used for training the secondary model, so as to obtain model weights corresponding to each model, so as to subsequently improve the accuracy of pollen concentration prediction results.

[0086] Among them, the model weight is the weight value of the corresponding first-level model. Since the results corresponding to different algorithm models may differ from the actual results, the accuracy of the prediction results of different models can be determined based on multi-layer training. The higher the accuracy of the output result, the larger the corresponding weight. Therefore, the model weights corresponding to each model can be obtained through training, and multi-model joint prediction can be realized to improve the accuracy of pollen forecast results.

[0087] S14, configuring the model weights to the corresponding primary models to generate a pollen prediction model.

[0088] It can be understood that the model weights at the training site are assigned to each primary model to obtain a pollen forecasting model so that when real-time data is subsequently input, fast and accurate predictions can be made.

[0089] S2, receiving multiple allergy factors set by the user end, obtaining real-time data of the corresponding display area based on the allergy factors, inputting the real-time data into the pollen prediction model to obtain pollen concentration data, inputting the pollen concentration data into the pollen health impact model for calculation, and obtaining the pollen impact logarithm.

[0090] It should be noted that since different people may be allergic to different pollens to different degrees, in order to provide targeted pollen forecasts and warnings of health impacts for each person, the allergy factors set by the user can be received for subsequent customized health assessments of the impact of pollen on the person. That is, users can input specific allergy factors based on their own allergy history and medical advice, so that the system can provide targeted forecasts and health impact assessments based on these factors.

[0091] It can be understood that the user end is the information terminal of the person who checks the pollen forecast, which can be a mobile phone, computer, etc. The allergy factor is the specific allergy-causing factor set by the user, such as rapeseed pollen, cherry blossom pollen, etc. The real-time data is the latest data of the environmental conditions in the current display area, such as temperature, humidity, wind speed, etc. These factors directly affect the spread and concentration of pollen. The system will continuously obtain real-time environmental data of the display area to ensure that the pollen forecast model uses the latest information for prediction, thereby improving the accuracy of the prediction.

[0092] Among them, the pollen concentration data is the predicted value of the current pollen concentration calculated by the pollen forecast model, the pollen health impact model is a model that evaluates the potential impact of pollen concentration data on user health, and the pollen impact logarithm is a value calculated based on the health impact model, which reflects the potential impact of pollen concentration on user health.

[0093] It is not difficult to understand that users can set their own sensitive allergy factors, such as a certain type of pollen. The system will obtain real-time data of the display area based on these factors, input these real-time data into the pollen forecast model, obtain the current pollen concentration, and then input the pollen concentration data into the pollen health impact model to calculate the impact of pollen on health.

[0094] Through the above-mentioned implementation manner, the present invention can obtain real-time data of user allergy factors and display areas, accurately predict pollen concentrations, and evaluate their impact on health, thereby providing users with timely and personalized health risk warnings.

[0095] In some embodiments, in step S2 (receiving multiple allergy factors set by the user terminal, obtaining real-time data of the corresponding display area based on the allergy factors, inputting the real-time data into the pollen prediction model to obtain pollen concentration data, and inputting the pollen concentration data into the pollen health impact model for calculation to obtain the pollen impact logarithm), specific implementation methods include:

[0096] S21, receiving a plurality of allergy factors set by a user terminal, and acquiring real-time data of the corresponding display area based on the allergy factors.

[0097] It is understandable that real-time data in the corresponding display area is obtained according to the allergy factor so that the pollen concentration corresponding to the allergy factor can be accurately predicted, thereby improving the accuracy of the pollen impact logarithm so as to provide allergy prompts to users in a timely manner.

[0098] S22, inputting the real-time data into the pollen prediction model to obtain the real-time prediction results corresponding to each of the first-level models, and obtaining the pollen concentration data corresponding to the display area according to the real-time prediction results and the corresponding model weights.

[0099] It can be understood that the real-time data is input into the pollen forecast model to obtain the real-time prediction results of the corresponding predictions of each first-level model, and then the pollen concentration data corresponding to the display area is obtained according to the real-time prediction results and the model weights, so as to be subsequently input into the pollen health impact model to calculate a more accurate pollen impact logarithm, and then, timely provide personnel with pollen display and health impact reminders.

[0100] Among them, the real-time prediction results are the data results predicted by the first-level model after analyzing the real-time data, and the pollen concentration data are the pollen concentration results predicted by the pollen forecasting model.

[0101] S23, inputting the pollen concentration data into a pollen health impact model for calculation to obtain a pollen impact logarithm.

[0102] It can be understood that by inputting pollen concentration data into the pollen health impact model, the corresponding pollen impact logarithm can be quickly calculated to obtain the degree of impact of the corresponding allergic factors in the current display area on human health, so that personnel can take corresponding measures in time to reduce the risk of allergies.

[0103] In some embodiments, in step S23 (the step of inputting the pollen concentration data into the pollen health impact model for calculation to obtain the pollen impact logarithm), the specific implementation method includes:

[0104] S231, the pollen impact logarithm is obtained by the following formula,

[0105] ;

[0106] in, is the logarithm of pollen impact, For the The number of outpatient visits per day, For the The expected number of outpatient visits per day, is the intercept, is the period dummy variable, is a nonparametric smoothing spline function, is a nonparametric plate spline function, For the date, is the pollen concentration, is the collaborative meteorological element variable, is a non-coordinated meteorological element variable, is the pollutant concentration, For degrees of freedom.

[0107] Understandably, the use of natural logarithm transformation is a common practice in statistical and epidemiological models, which can deal with skewed distributions of dependent variables, make models more stable, and help explain nonlinear relationships between variables.

[0108] in, Represents the expected number of outpatient visits on observation day K, for example, the expected number of outpatient visits on September 1. This is a predicted value or theoretical average value, which is used to represent the expected number of outpatient visits under given conditions (such as pollen concentration, meteorological conditions, etc.). For the The expected value of the number of outpatient visits per day is input into the logarithmic value of the logarithmic function. The larger the output logarithmic value, the closer the exposure-response relationship between pollen and the number of outpatient visits is. That is, the model establishes an exposure-response relationship between pollen and the number of outpatient visits, which means that the higher the degree of influence between the two, the greater the exposure-response relationship output by this model. This exposure-response relationship is a key parameter for judging the impact of pollen on human health. is a nonparametric smoothing spline function and The non-parametric flat spline functions are all existing calculation functions. The degree of freedom is usually determined by cross-validation and other methods. That is, when using spline function for non-parametric smoothing, the degree of freedom is an important parameter that controls the complexity of smoothing and the flexibility of fitting. The higher the degree of freedom, the more complex the curve that the model can fit, and the lower the degree of freedom, the simpler the curve that the model fits.

[0109] It is worth mentioning that The numerical value involves the debugging process of the model, which generally requires Assign a value of about 1-100, then edit the statement to loop and calculate the AIC value (goodness of model establishment) under different degrees of freedom. The model will take the value of the degree of freedom corresponding to the smallest AIC value (the most accurate model simulation) as the final degree of freedom. It can also be preset manually. Among them, AIC is the Akaike Information Criterion, a statistical method for model selection, which provides a standard for measuring the goodness of fit of statistical models. AIC not only considers the degree of fit of the model to the data (usually achieved by maximizing the likelihood function), but also considers the number of parameters in the model, thereby penalizing the complexity of the model. Therefore, the smaller the AIC value, it usually means that the model maintains a lower complexity while fitting the data, that is, the model is more concise and has stronger predictive ability.

[0110] Among them, the coordinated meteorological element variables Including meteorological elements such as temperature, wind speed and relative humidity, refers to the date The corresponding values ​​of temperature, wind speed, etc., among which the synergistic meteorological element variable refers to the synergistic relationship between pollen and temperature on human health. The model is a medical model. For example, rising temperature may increase the amount of air inhaled by the human body, increase blood viscosity, and other medical changes, which will be related to human allergies and may aggravate simple allergies. It is a non-synergistic meteorological element variable. A non-synergistic meteorological element variable means that pollen and the corresponding conditions have no effect on human health and do not have a synergistic relationship. A non-synergistic meteorological element variable refers to a meteorological factor (for example, air pressure, etc.) that will not have a synergistic effect on allergies with pollen from a medical mechanism. The judgment standard is determined based on actual medical evaluation and can also be artificially set in advance.

[0111] It is not difficult to understand that the more people visit the allergy clinic on the corresponding date, the greater the impact of the current pollen concentration on people's health. To calculate the intercept corresponding to the model, it can be manually preset, the period dummy variable It is a 1-7 day cycle variable that is preset manually. It is a common variable used to consider the impact of different days of the week on the number of outpatient visits or other similar indicators. This is because people's activity patterns, health conditions, and medical behavior may vary depending on the day of the week. For example, the specific value of pollutant concentration is the concentration value of PM2.5, NO2, O3, SO2, etc. obtained by sensors.

[0112] For example, people may be more inclined to rest or do outdoor activities on weekends, while they are more busy at work on weekdays. These differences may affect the number of outpatient visits. In epidemiological models or any models that need to consider time factors, adding DOW variables can help the model capture these cyclical changes more accurately, thereby improving the model's prediction accuracy. By training the model, the coefficient (or effect) of DOW will be automatically adjusted to reflect the specific impact of different days of the week on the expected number of outpatient visits.

[0113] S232, sending the pollen impact logarithm to the management end.

[0114] It is understandable that the pollen impact logarithm is sent to the management end so that the management personnel can check it in time, thereby making forecast reminders and reducing allergic reactions of personnel.

[0115] Among them, the management end is the information terminal of the person who manages the pollen forecast broadcast.

[0116] S3, when it is determined that the pollen influence logarithm is greater than a preset influence value, based on the allergen display strategy and the planting areas of each allergen factor in the display area, generating an initial diffusion area corresponding to each planting area in the display area.

[0117] It is understandable that when pollen reaches a certain concentration, it may have a health impact on people. Therefore, the initial diffusion area can be determined based on the pollen impact logarithm, so that the real-time location of the people can be obtained for pollen allergy warnings, high-risk areas can be preliminarily identified, and basic data support can be provided for subsequent risk management and response measures.

[0118] Among them, the preset impact value is a preset baseline pollen value that affects health, the planting area is the area within the display area that corresponds to the planting allergy factor, such as the rapeseed planting area and the cherry blossom planting area, and the initial diffusion area is an area that may be affected by pollen based on the planting area and display strategy.

[0119] It is not difficult to understand that when the calculated pollen impact logarithm exceeds the preset health risk value, the system will generate an initial diffusion area based on the planting conditions and display strategies of different allergy factors in the display area. This area is a potential high-risk area.

[0120] In some embodiments, the specific implementation of step S3 (when determining that the pollen influence logarithm is greater than a preset influence value, generating the initial diffusion area corresponding to each planting area in the display area based on the allergen display strategy and the planting area of ​​each allergen in the display area) includes:

[0121] S31, when it is determined that the pollen influence logarithm is greater than a preset influence value, the planting areas of each allergic factor in the display area and the sub-planting areas corresponding to each planting area are obtained.

[0122] It is understandable that when the pollen influence logarithm is greater than the preset influence value, it means that the corresponding pollen will spread. Therefore, it is necessary to obtain the planting area and sub-planting area of ​​the corresponding allergy factor to facilitate the subsequent calculation of the diffusion area of ​​the corresponding allergy factor and display prompts to personnel.

[0123] Among them, the sub-planting area is the area of ​​the planting area corresponding to each type of flower.

[0124] S32, obtaining the total planting area corresponding to the display area, and obtaining the planting proportion corresponding to each of the allergic factors according to the ratio of the sub-planting area to the total planting area.

[0125] It is understandable that in the same display area, there may be planting areas corresponding to multiple allergic factors, and the same allergic factor may have multiple planting areas. Therefore, in order to calculate the concentration data of the corresponding planting area and thus determine the corresponding initial diffusion area, the planting proportion of each allergic factor can be obtained to improve the accuracy of the subsequent initial diffusion area.

[0126] Among them, the total planting area is the sum of the planting areas of all allergens in the display area, and the planting proportion is the proportion of each planting area corresponding to each allergen, that is, the ratio of the sub-planting area to the total planting area.

[0127] S33, obtaining sub-pollen concentration data corresponding to each of the planting areas based on the product of the planting proportion and the corresponding pollen concentration data.

[0128] It can be understood that by multiplying the planting proportion and the pollen concentration data of the corresponding display area, the sub-pollen concentration data corresponding to each planting area can be obtained, so that the corresponding theoretical expansion distance can be determined through the sub-pollen concentration data, and then the initial diffusion area can be obtained.

[0129] Among them, the sub-pollen concentration data is the pollen concentration corresponding to each planting area, that is, it can be obtained by multiplying the planting ratio and the corresponding pollen concentration data.

[0130] S34, obtaining a concentration diffusion coefficient according to a ratio of the sub-pollen concentration data to the preset area concentration data, and obtaining a theoretical diffusion distance corresponding to the planting area based on a product of the concentration diffusion coefficient and the preset diffusion distance.

[0131] It is understandable that the degree of diffusion corresponding to different pollen concentrations will also have certain differences. When there is no interference from the external environment, the higher the concentration, the larger the corresponding diffusion area. Therefore, in order to obtain an accurate diffusion distance, the sub-pollen concentration data can be compared with the preset area concentration data, and then the corresponding concentration diffusion coefficient can be obtained, so that the theoretical diffusion distance can be calculated, and further the corresponding initial diffusion area can be determined later.

[0132] Among them, the preset area concentration data is the preset pollen concentration baseline value, the concentration diffusion coefficient is the proportional coefficient value of the pollen concentration diffusion, that is, the ratio of the sub-pollen concentration data to the preset area concentration data, the preset diffusion distance is the preset diffusion distance, and the theoretical diffusion distance is the diffusion distance corresponding to the current pollen concentration, that is, the product of the concentration diffusion coefficient and the preset diffusion distance.

[0133] S35, performing movement processing on the area outline pixel points of the planting area based on the theoretical diffusion distance to generate an initial diffusion area corresponding to each planting area in the display area.

[0134] It can be understood that by moving the area outline pixel points of the planting area outward by the theoretical diffusion distance, the moved pixel points can be connected, thereby determining the initial diffusion area corresponding to each planting area.

[0135] The regional contour pixel points are the contour pixel points of the planting area image.

[0136] In some embodiments, the specific implementation of step S35 (moving the pixel points of the area outline of the planting area based on the theoretical diffusion distance to generate the initial diffusion area corresponding to each planting area in the display area) includes:

[0137] S351, selecting any pixel point in the display area as a coordinate origin to construct a display coordinate system, and performing coordinate processing on the display area based on the display coordinate system.

[0138] It is understandable that in order to facilitate the subsequent determination of the position of the pixel point after the movement, thereby determining the initial diffusion area, a display coordinate system can be constructed in the display area to coordinate the display area.

[0139] The display coordinate system is a coordinate system constructed for the display area.

[0140] S352, obtaining the contour pixel points of each planting area in the display area as the regional contour pixel points, and determining the regional center coordinates of the planting area and the regional center point corresponding to the regional center coordinates according to the maximum horizontal coordinate, the minimum horizontal coordinate, the maximum vertical coordinate and the minimum vertical coordinate of the regional contour pixel points.

[0141] It can be understood that the regional contour pixel points are the contour pixel points of each planting area in the display area, the maximum horizontal coordinate is the maximum value of the horizontal coordinate corresponding to the regional contour pixel points, the minimum horizontal coordinate is the minimum value of the horizontal coordinate corresponding to the regional contour pixel points, the maximum vertical coordinate is the maximum value of the vertical coordinate corresponding to the regional contour pixel points, the minimum vertical coordinate is the minimum value of the vertical coordinate corresponding to the regional contour pixel points, the regional center coordinates are the coordinates of the center point of the planting area, and the regional center point is the center point of the planting area.

[0142] It is not difficult to understand that the existing image processing technology OpenCV can be used to extract the contour pixel points in the planting area, and then determine the corresponding coordinates, and average the horizontal and vertical coordinates to obtain the center coordinates of the area, thereby determining the center point of the area, so as to facilitate the subsequent movement of the area contour pixel points and determine the initial diffusion area.

[0143] S353, sequentially connecting the area center point with the area outline pixel points of the corresponding planting area to obtain a diffusion connection line.

[0144] It can be understood that the diffusion connection line is a connection line where the pixel points diffuse outward, that is, a connection line between the center point of the region and the pixel points of the region outline corresponding to the planting region.

[0145] Through the above implementation, the present invention can obtain the diffusion connection line, so as to determine the accurate moving direction of the pixel point later.

[0146] S354: Moving the region outline pixel points along the diffusion connection line and away from the region center point based on the theoretical diffusion distance to obtain diffused region outline pixel points.

[0147] It can be understood that in order to determine the area corresponding to the diffusion of pollen, the regional contour pixels can be moved along the diffusion connection line away from the center point of the area by the theoretical diffusion distance, and the position of the pixels after the move can be determined, so that the corresponding regional contour pixels after diffusion can be connected subsequently to obtain the initial diffusion area.

[0148] S355, connecting adjacent pixel points of the diffused area contours to obtain initial diffusion areas corresponding to the planting areas.

[0149] It can be understood that by sequentially connecting adjacent pixel points of the diffused area contours, the initial diffusion areas corresponding to the various planting areas circled within the connecting lines can be obtained, so as to provide subsequent display prompts to personnel and reduce the risk of pollen allergies.

[0150] It is worth mentioning that wind direction will also have a certain impact on the pollen diffusion area. Therefore, the corresponding wind direction can be obtained, so that the pollen diffusion area can be adjusted according to the actual wind direction. That is, the regional contour pixel points corresponding to the planting area will be moved farther in the wind direction, and the moving distance of the regional contour points corresponding to the opposite direction of the wind direction will be shorter, so as to obtain the actual pollen diffusion area for subsequent intuitive display to personnel.

[0151] It should be noted that, since the corresponding pollen concentration will change during the pollen diffusion process, different pollen diffusion areas with different concentrations can be displayed differently through different pixel values, so that people can intuitively view the pollen concentration of the corresponding area, thereby timely warning allergic people, therefore, it also includes:

[0152] A1. Determine a concentration pixel value corresponding to the sub-pollen concentration data according to the sub-pollen concentration data and a preset concentration comparison table, wherein the preset concentration comparison table has a one-to-one correspondence between preset concentration intervals and concentration pixel values.

[0153] It is understandable that the concentration pixel value corresponding to the sub-pollen concentration data is found in the preset concentration comparison table so as to be subsequently filled in the corresponding planting area to intuitively display the pollen concentration information of the planting area.

[0154] Among them, the preset concentration comparison table includes preset concentration areas and concentration pixel values ​​corresponding to preset concentration intervals. It is pre-set, and the higher the pollen concentration corresponding to the preset concentration interval, the darker the corresponding concentration pixel value is displayed. The preset concentration interval is a pre-set pollen concentration interval, and the concentration pixel value is a pixel value representing the pollen concentration, such as dark red, rose red, light red, etc.

[0155] A2, filling the concentration pixel value into the corresponding planting area, and obtaining a gradient area according to the difference set between the initial diffusion area and the planting area.

[0156] It can be understood that the corresponding concentration pixel values ​​are filled in the corresponding planting area, and the gradient area is obtained according to the difference between the initial diffusion area and the planting area, so that the gradient area can be filled with corresponding colors for display later.

[0157] Among them, the gradient area is the area where the pollen concentration gradually decreases with the diffusion concentration.

[0158] A3, sequentially moving the region contour pixel points along the diffusion connection line and away from the region center point based on a preset gradient distance to obtain gradient contour pixel points.

[0159] It is understandable that, as pollen diffuses, the corresponding pollen concentration varies with the diffusion distance. Therefore, the regional contour pixels can be moved in sequence along the diffusion connection line away from the center point of the region according to the preset gradient distance, thereby determining the golden-edged contour pixels, so as to subsequently obtain gradient sub-regions of different concentrations.

[0160] The preset gradient distance is a preset gradient distance of different concentrations, and the gradient outline pixel points are pixel points that determine the outline of the gradient area, that is, pixel points after moving the preset gradient distance.

[0161] A4, connecting adjacent gradient contour pixel points to obtain gradient contour lines in a gradient region, segmenting the gradient region according to the gradient contour lines to obtain a plurality of gradient sub-regions, and sorting the gradient sub-regions along the diffusion connection lines and in a direction away from the center point of the region to obtain a gradient sub-sequence.

[0162] It can be understood that by connecting adjacent gradient contour pixel points to obtain multiple gradient contour lines in the gradient area, the gradient area can be segmented, and multiple gradient sub-areas can be obtained. The gradient sub-areas can then be sorted along the diffusion connection lines and in the direction away from the center point of the area to obtain a gradient sub-sequence. For example, the gradient sub-sequence can be (gradient sub-area 1, gradient sub-area 2, gradient sub-area 3), so that the corresponding pixel values ​​can be filled in the gradient sub-areas in turn according to the gradient sub-sequence for intuitive display.

[0163] The gradient contour line is a line connecting adjacent gradient contour pixels, the gradient sub-region is a region after the gradient region is segmented, and the gradient sub-sequence is a region sequence after the gradient sub-regions are arranged.

[0164] A5, based on the ratio of the theoretical diffusion distance to the gradient distance, the gradient times are obtained.

[0165] It is understandable that in order to determine the pixel value corresponding to the gradient sub-region, the gradient times can be obtained by the ratio of the theoretical diffusion distance and the gradient distance, so as to subsequently determine the number of pixel values ​​to fill and display the gradient sub-region.

[0166] The number of gradients refers to the number of times the pollen concentration gradients.

[0167] A6, according to the number of gradients, the concentration pixel value is sequentially increased according to a preset attenuation value to obtain a plurality of gradient pixel values, and the gradient pixel values ​​are sequentially filled into the gradient sub-regions in the gradient sub-sequence.

[0168] It can be understood that the color value range corresponding to the pixel value is 0 to 255, black is 0 and white is 255. Therefore, as the value increases, the corresponding color depth becomes lighter. Therefore, the concentration pixel value can be increased in sequence according to the preset attenuation value based on the number of gradients, so as to determine the gradient pixel value corresponding to each gradient sub-region, so as to distinguish the corresponding gradient sub-region.

[0169] For example, Figure 2 As shown, the gradient pixel values ​​are filled into the corresponding gradient sub-areas in sequence, so as to be displayed to personnel for intuitive viewing later.

[0170] The preset attenuation value is a preset value corresponding to the attenuation of the color as the concentration decreases, and the gradient pixel value is a pixel value corresponding to the gradient sub-area.

[0171] It should be noted that since the distances between multiple planting areas may be relatively close, when pollen spreads, the initial diffusion areas of different planting areas may overlap. Therefore, when displaying, the pixel values ​​of the corresponding gradient areas need to be changed to provide intuitive display prompts to personnel. Therefore, it also includes:

[0172] B1, when it is determined that the initial diffusion area has an overlapping area, the corresponding overlapping area is used as a fused diffusion area, and the corresponding initial diffusion area is used as an influencing diffusion area, and the sub-pollen concentration data of the influencing diffusion area is obtained as the influencing concentration data.

[0173] It can be understood that the overlapping area is the area where multiple initial diffusion areas overlap, the fused diffusion area is the area where pollen fuses during the diffusion process, that is, the overlapping area, the affected diffusion area is the corresponding initial diffusion area, and the affected concentration data is the sub-pollen concentration data of the affected diffusion area.

[0174] For example: Figure 3 As shown, the fused diffusion area is determined by the initial diffusion area, so that the corresponding fused pixel value can be subsequently determined and filled in the fused diffusion area, so that personnel can intuitively view the pollen concentration information of the corresponding area and take timely protection.

[0175] Through the above implementation, the present invention can determine the fused diffusion area, the affected diffusion area and the affected concentration data, so as to subsequently determine the corresponding fused pixel value, and then fill it in the corresponding area for display.

[0176] B2, obtaining an independent diffusion area according to the difference between the affected diffusion area and the corresponding overlapping area, and using the affected concentration data corresponding to the affected diffusion area as the independent pollen concentration data of the independent diffusion area.

[0177] It can be understood that the independent diffusion area is the area within the influence diffusion area excluding the overlapping part, that is, the difference between the influence diffusion area and the corresponding overlapping area, and the independent pollen concentration data is the pollen concentration data of the independent diffusion area, that is, the influence concentration data corresponding to the influence diffusion area.

[0178] B3, obtaining a plurality of regional pixel values ​​corresponding to the fused diffusion region, and a regional area corresponding to each of the regional pixel values.

[0179] It can be understood that because the fused diffusion area is an overlapping area of ​​multiple initial diffusion areas, which may contain multiple gradient sub-areas, it has multiple corresponding regional pixel values. At the same time, the regional area corresponding to each regional pixel value can be obtained, so as to subsequently determine the fused pixel value of the corresponding fused diffusion area.

[0180] The regional pixel value is the pixel value filled in each region, and the regional area is the area occupied by the pixel value of each region in the fusion diffusion region.

[0181] B4, calculating the ratio of the area of ​​each region corresponding to the pixel value of each region to obtain the area ratio of the pixel value of the region.

[0182] It can be understood that the area ratio is the ratio of the area corresponding to the area pixel value. For example, when there are three area pixel values, and the corresponding areas are 2 , 4 and 6 When , the corresponding area ratio is 1:2:3.

[0183] Through the above implementation, the corresponding area ratio is calculated so that the corresponding fused pixel value can be obtained later through the area ratio.

[0184] B5. Perform calculation based on the area proportion and the corresponding pixel value of the region to obtain a fused pixel value corresponding to the fused diffusion region.

[0185] It can be understood that the RGB values ​​in the regional pixel values ​​are respectively calculated as the mean of R value, the mean of G value and the mean of B value according to the area ratio. For example, when the area ratio is 1:2:3, the corresponding regional pixel values ​​are (0,0,0), (125,88,100) and (255,122,124). = =169, = =90, = =95, then the corresponding fused pixel value can be obtained as (169,90,95).

[0186] Among them, the fused pixel value is the pixel value corresponding to the fused diffusion area, and then the corresponding fused diffusion area can be filled with the corresponding color for display.

[0187] S4, receiving the regional data of the display area, customizing and adjusting the initial diffusion area according to the regional data to obtain the allergen area corresponding to the display area, generating warning information based on the real-time location information of the user terminal and the allergen area and sending it to the user terminal.

[0188] It should be noted that in order to make specific adjustments to the initial diffusion area based on actual conditions, so as to make it more in line with actual conditions and ensure that users can obtain risk warnings about allergens in a timely manner and take corresponding protective measures, it is possible to receive regional data of the display area, customize the initial diffusion area, and generate early warning information and send it to the user end, so that personnel can intuitively view the corresponding affected area and take protective measures.

[0189] It is understandable that the server first receives the specific data of the display area, and then adjusts the initial diffusion area generated in the previous step based on these data. The high-risk area finally determined will generate warning information based on the user's real-time location and send it to the user.

[0190] Among them, regional data is the specific geographical and environmental data in the display area, the allergen area is the final high-risk area after adjustment, the real-time location information is the user's current location, and the warning information is the warning information generated based on the real-time location information and the high-risk area.

[0191] Through the above implementation, the present invention can adjust the initial diffusion area to ensure that the high-risk area is more accurate. Combined with the user's real-time location information, early warning information is generated to ensure that the user can get accurate risk warnings in time and improve the protection effect.

[0192] In some embodiments, the specific implementation of step S4 (receiving the regional data of the display area, customizing and adjusting the initial diffusion area according to the regional data, obtaining the allergen area corresponding to the display area, and generating warning information based on the real-time location information of the user terminal and the allergen area and sending it to the user terminal) includes:

[0193] S41, receiving regional data of the display area, wherein the regional data includes temperature, rainfall and wind speed.

[0194] It can be understood that the regional data are the data that affect pollen diffusion in the corresponding display data, including temperature, rainfall and wind speed.

[0195] It is not difficult to understand that when the temperature is higher, the corresponding molecules in the air are more active, which will assist the spread of pollen, and when there is rainfall in the display area, it will inhibit the spread of pollen. At the same time, the greater the wind speed, the greater the pollen will be assisted to spread to a farther distance, thereby obtaining the temperature, rainfall and wind speed, so as to obtain the accurate allergen area later.

[0196] S42, adjusting the theoretical diffusion distance according to the temperature, the rainfall and the wind speed to obtain an actual diffusion distance.

[0197] It can be understood that the theoretical diffusion distance is adjusted by displaying the temperature, rainfall and wind speed corresponding to the area, thereby obtaining the actual diffusion distance.

[0198] The actual diffusion distance is the actual distance that pollen diffuses in the display area.

[0199] The corresponding actual diffusion distance is calculated by the following formula:

[0200]

[0201] in, is the actual diffusion distance, is the theoretical diffusion distance, is the actual temperature, For the preset temperature, is the actual wind speed, For the preset wind speed, is the actual rainfall, To preset rainfall, is the temperature weight, is the wind speed weight, is the rainfall weight.

[0202] It is understandable that when the actual temperature The higher the actual wind speed The larger the value, the corresponding actual rainfall The smaller the temperature, the larger the actual diffusion distance will be, and the corresponding pollen diffusion distance will become farther. The lower the actual wind speed The smaller the corresponding actual rainfall The larger it is, the less pollen can spread, and the actual diffusion distance will also decrease.

[0203] Among them, the preset temperature is a preset temperature reference value, the preset wind speed is a preset wind speed reference value, the preset rainfall is a preset rainfall reference value, and the preset temperature, preset wind speed and preset rainfall are coefficient values ​​used to determine the adjustment of the theoretical diffusion distance in order to obtain an accurate actual diffusion distance, thereby accurately providing allergen area prompts, allowing personnel to take timely protection and reduce the risk of allergies.

[0204] It is not difficult to understand that the temperature weight, wind speed weight and rainfall weight are artificially set in advance according to the influence of the corresponding factors on pollen diffusion, so as to improve the accuracy of the actual diffusion distance.

[0205] The increased adjusted temperature weight, increased adjusted wind speed weight, and decreased adjusted rainfall weight are obtained through the following formula.

[0206]

[0207] in, To increase the weight of the adjusted temperature, To adjust the diffusion distance, is the deviation adjustment value, z is the weight ratio, To increase the adjusted wind speed weight, To reduce the adjusted rainfall weight.

[0208] It can be understood that when the artificially input adjustment expansion distance is larger than the calculated diffusion distance, it means that the actual diffusion distance of the current output is smaller than the diffusion distance in the real environment. At the same time, it indicates that the corresponding weight deviation is large, so adjustment is needed, that is, the temperature weight and wind speed weight can be increased to increase the diffusion distance, while the rainfall weight can be reduced to increase the diffusion distance.

[0209] Among them, the difference between the manually input adjusted diffusion distance and the calculated actual diffusion distance can be compared with the deviation adjustment value and multiplied by the weight ratio, so that the weight can be adjusted according to the distance difference in the actual situation, thereby improving the accuracy of the adjusted weight. At the same time, the weight ratio is added so that the current weight value can be finely adjusted in proportion to further improve the accuracy of the weight adjustment, so that the actual diffusion distance data trained in the future will be more accurate, thereby improving the accuracy of the forecast.

[0210] In addition, the R deviation adjustment value is the reference value for weight deviation adjustment, which is preset manually. z is determined according to the weight value of the corresponding category, that is, when the temperature weight is adjusted, the corresponding z is and ratio, and is the comprehensive weight, which is the temperature weight of the three weights , wind speed weight and rainfall weight When the wind speed weight is adjusted, the corresponding z is and When the rainfall weight is adjusted, the corresponding z is and ratio.

[0211] It is not difficult to understand that the greater the difference between the adjusted diffusion distance and the output actual diffusion distance, the greater the data deviation of the current temperature weight, wind speed weight and rainfall weight. The greater the adjusted temperature weight, the greater the adjusted wind speed weight, and the smaller the adjusted rainfall weight, so that the trained weight value can predict a more accurate actual diffusion distance.

[0212] The reduced adjusted temperature weight, reduced adjusted wind speed weight and increased adjusted rainfall weight are obtained through the following formula.

[0213]

[0214] in, To reduce the adjusted temperature weight, To adjust the diffusion distance, is the deviation adjustment value, z is the weight ratio, To reduce the adjusted wind speed weight, To increase the adjusted rainfall weight.

[0215] It can be understood that when the artificially input adjustment expansion distance is smaller than the calculated diffusion distance, it means that the actual diffusion distance of the current output is larger than the diffusion distance in the real environment, and it also indicates that the corresponding weight deviation is large, so adjustment is needed, that is, the temperature weight and wind speed weight can be reduced to reduce the diffusion distance, and the rainfall weight can be increased to reduce the diffusion distance.

[0216] Among them, the difference between the manually input adjusted diffusion distance and the calculated actual diffusion distance can be compared with the deviation adjustment value and multiplied by the weight ratio, so that the weight can be adjusted according to the distance difference in the actual situation, thereby improving the accuracy of the adjusted weight. At the same time, the weight ratio is added to further improve the accuracy of the weight adjustment, so that the actual diffusion distance data trained later is more accurate, thereby improving the accuracy of the forecast.

[0217] In addition, the R deviation adjustment value is the reference value for weight deviation adjustment, which is preset manually. z is determined according to the weight value of the corresponding category, that is, when the temperature weight is adjusted, the corresponding z is and ratio, and is the comprehensive weight, which is the temperature weight of the three weights , wind speed weight and rainfall weight When the wind speed weight is adjusted, the corresponding z is and When the rainfall weight is adjusted, the corresponding z is and ratio.

[0218] It is not difficult to understand that the greater the difference between the adjusted diffusion distance and the output actual diffusion distance, the greater the data deviation of the current temperature weight, wind speed weight and rainfall weight, and the smaller the adjusted temperature weight and wind speed weight will be, and the larger the adjusted rainfall weight will be, so that the trained weight value can predict a more accurate actual diffusion distance.

[0219] S43, performing movement processing on the area outline pixel points of the planting area based on the actual diffusion distance to generate an allergen area corresponding to the display area.

[0220] It can be understood that by moving the regional contour pixel points of the planting area along the diffusion connection line away from the regional center point by the actual diffusion distance, the corresponding pixel point position can be determined, and the adjacent pixel points are connected to obtain the allergen contour line, wherein the area enclosed by the allergen connection line is the allergen area.

[0221] The allergen contour is the contour of the allergen area.

[0222] S44, acquiring the real-time location information of the user terminal, determining that the real-time location information is in the allergen area, generating warning information and sending it to the user terminal.

[0223] It can be understood that by obtaining the real-time location information of the user end, the location of the corresponding person can be determined in time. When the real-time location information is in the allergen area, it means that the user is in the corresponding pollen diffusion area and has an allergy risk. Therefore, an early warning prompt message will be generated and sent to the user end, so that the person can stay away in time or take corresponding protective measures to reduce the risk of allergies.

[0224] The real-time location information is the real-time geographical location information corresponding to the user end, and the warning information is the information to warn people about pollen allergies.

[0225] See also Figure 4 , is a schematic diagram of a structure of a prediction and management system for pollen multi-dimensional information provided by an embodiment of the present invention, the prediction and management system for pollen multi-dimensional information includes:

[0226] The training module is used to statistically display the historical project data of the area as the initial training set, input the initial training set into multiple first-level models in the initial forecast model for training to obtain intermediate prediction results and combine them, and obtain the intermediate training set and input it into the second-level model for training to generate a pollen forecast model.

[0227] The calculation module is used to receive multiple allergy factors set by the user end, obtain real-time data of the corresponding display area based on the allergy factors, input the real-time data into the pollen prediction model to obtain pollen concentration data, input the pollen concentration data into the pollen health impact model for calculation, and obtain the pollen impact logarithm.

[0228] A generation module is used to generate an initial diffusion area corresponding to each planting area in the display area based on the allergen display strategy and the planting area of ​​each allergic factor in the display area when it is determined that the pollen influence logarithm is greater than a preset influence value.

[0229] The adjustment module is used to receive the regional data of the display area, customize the initial diffusion area according to the regional data, obtain the allergen area corresponding to the display area, and generate warning information based on the real-time location information of the user terminal and the allergen area and send it to the user terminal.

[0230] See also Figure 5 , is a schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention, the electronic device 50 includes: a processor 51, a memory 52 and a computer program; wherein

[0231] The memory 52 is used to store the computer program, which may also be a flash memory. The computer program is, for example, an application program, a functional module, etc. for implementing the above method.

[0232] The processor 51 is used to execute the computer program stored in the memory to implement each step performed by the device in the above method. For details, please refer to the relevant description in the above method embodiment.

[0233] Optionally, the memory 52 may be independent or integrated with the processor 51 .

[0234] When the memory 52 is a device independent of the processor 51, the device may further include:

[0235] The bus 53 is used to connect the memory 52 and the processor 51 .

[0236] The present invention also provides a readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, it is used to implement the methods provided by the various embodiments described above.

[0237] Among them, the readable storage medium can be a computer storage medium or a communication medium. The communication medium includes any medium that facilitates the transmission of a computer program from one place to another. The computer storage medium can be any available medium that can be accessed by a general-purpose or special-purpose computer. For example, a readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application-specific integrated circuit (Application Specific Integrated Circuits, referred to as: ASIC). In addition, the ASIC can be located in a user device. Of course, the processor and the readable storage medium can also exist in a communication device as discrete components. The readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0238] The present invention also provides a program product, which includes an execution instruction, which is stored in a readable storage medium. At least one processor of a device can read the execution instruction from the readable storage medium, and at least one processor executes the execution instruction so that the device implements the methods provided in the above various embodiments.

[0239] In the embodiments of the above-mentioned devices, it should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the present invention may be directly implemented as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0240] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting and managing multi-dimensional pollen information, characterized in that: include: The historical project data of the statistical display area is used as an initial training set, and the initial training set is input into multiple primary models in the initial forecast model for training to obtain intermediate prediction results and combine them, and the intermediate training set is input into the secondary model for training to generate a pollen forecast model; Receive multiple allergy factors set by the user end, obtain real-time data of the corresponding display area based on the allergy factors, input the real-time data into a pollen prediction model to obtain pollen concentration data, input the pollen concentration data into a pollen health impact model for calculation, and obtain the pollen impact logarithm; When it is determined that the pollen impact logarithm is greater than the preset impact value, based on the allergen display strategy and the planting areas of each allergen in the display area, an initial diffusion area corresponding to each planting area in the display area is generated; receiving the regional data of the display area, customizing and adjusting the initial diffusion area according to the regional data to obtain the allergen area corresponding to the display area, and generating warning information based on the real-time location information of the user terminal and the allergen area and sending it to the user terminal; When it is determined that the pollen impact logarithm is greater than a preset impact value, based on the allergen display strategy and the planting areas of each allergen in the display area, generating an initial diffusion area corresponding to each planting area in the display area, including: When it is determined that the pollen influence logarithm is greater than the preset influence value, the planting areas of each allergic factor in the display area and the sub-planting areas corresponding to each planting area are obtained; The total planting area corresponding to the display area is obtained, and the planting proportion corresponding to each of the allergic factors is obtained according to the ratio of the sub-planting area to the total planting area; Based on the product of the planting proportion and the corresponding pollen concentration data, sub-pollen concentration data corresponding to each of the planting areas is obtained; According to the ratio of the sub-pollen concentration data and the preset area concentration data, a concentration diffusion coefficient is obtained, and based on the product of the concentration diffusion coefficient and the preset diffusion distance, a theoretical diffusion distance corresponding to the planting area is obtained; The pixel points of the area outline of the planting area are moved based on the theoretical diffusion distance to generate initial diffusion areas corresponding to each planting area in the display area.

2. The method according to claim 1, characterized in that The historical project data of the statistical display area is used as an initial training set, and the initial training set is input into multiple primary models in the initial forecast model for training to obtain intermediate prediction results and combine them, and the intermediate training set is input into the secondary model for training to generate a pollen forecast model, including: Counting historical project data of the display area as an initial training set, wherein the historical project data at least includes historical basic data and historical meteorological data; Inputting the initial training set into multiple first-level models in the initial forecast model for training, and obtaining intermediate prediction results corresponding to each of the first-level models; Combining the intermediate prediction results to obtain an intermediate training set, which is input into the secondary model for training, to obtain model weights corresponding to each of the primary models; The model weights are configured to the corresponding primary models to generate a pollen prediction model.

3. The method according to claim 2, characterized in that The receiving of multiple allergy factors set by the user terminal, obtaining real-time data of the corresponding display area based on the allergy factors, inputting the real-time data into the pollen prediction model to obtain pollen concentration data, and inputting the pollen concentration data into the pollen health impact model for calculation to obtain the pollen impact logarithm, including: Receiving multiple allergy factors set by a user terminal, and acquiring real-time data of the corresponding display area based on the allergy factors; Input the real-time data into the pollen prediction model to obtain the real-time prediction results corresponding to each of the first-level models, and obtain the pollen concentration data corresponding to the display area according to the real-time prediction results and the corresponding model weights; The pollen concentration data is input into the pollen health impact model for calculation to obtain the pollen impact logarithm.

4. The method according to claim 3, characterized in that The pollen concentration data is input into the pollen health impact model for calculation to obtain the pollen impact logarithm, including: The pollen impact logarithm is obtained by the following formula: ； in, is the logarithm of pollen impact, For the The number of outpatient visits per day, For the The expected number of outpatient visits per day, is the intercept, is the period dummy variable, is a nonparametric smoothing spline function, is a nonparametric plate spline function, For the date, is the pollen concentration, is the collaborative meteorological element variable, is a non-coordinated meteorological element variable, is the pollutant concentration, is the degree of freedom; The pollen impact logarithm is sent to a management terminal.

5. The method according to claim 1, characterized in that The moving processing of the area outline pixel points of the planting area based on the theoretical diffusion distance to generate the initial diffusion area corresponding to each planting area in the display area includes: Selecting any pixel point in the display area as a coordinate origin to construct a display coordinate system, and performing coordinate processing on the display area based on the display coordinate system; Acquire the outline pixel points of each planting area in the display area as the area outline pixel points, and determine the area center coordinates of the planting area and the area center point corresponding to the area center coordinates according to the maximum abscissa, minimum abscissa, maximum ordinate and minimum ordinate of the area outline pixel points; Sequentially connecting the center point of the region with the region outline pixel points of the corresponding planting region to obtain a diffusion connection line; Based on the theoretical diffusion distance, the region outline pixel points are moved along the diffusion connection line and in a direction away from the region center point to obtain diffused region outline pixel points; Adjacent pixels of the diffused area contours are connected to obtain initial diffusion areas corresponding to the planting areas.

6. The method according to claim 5, characterized in that Also includes: Determine a concentration pixel value corresponding to the sub-pollen concentration data according to the sub-pollen concentration data and a preset concentration comparison table, wherein the preset concentration comparison table has a one-to-one correspondence between preset concentration intervals and concentration pixel values; Filling the concentration pixel value into the corresponding planting area, and obtaining a gradient area according to the difference between the initial diffusion area and the planting area; Sequentially moving the region contour pixel points along the diffusion connection line and in a direction away from the region center point based on a preset gradient distance to obtain gradient contour pixel points; Connecting adjacent gradient contour pixel points to obtain gradient contour lines in a gradient region, segmenting the gradient region according to the gradient contour lines to obtain a plurality of gradient sub-regions, and sorting the gradient sub-regions along the diffusion connection line and in a direction away from the center point of the region to obtain a gradient sub-sequence; Based on the ratio of the theoretical diffusion distance to the gradient distance, the gradient times are obtained; The concentration pixel value is sequentially increased according to the preset attenuation value according to the number of gradients to obtain a plurality of gradient pixel values, and the gradient pixel values ​​are sequentially filled into the gradient sub-regions in the gradient sub-sequence.

7. The method according to claim 6, characterized in that Also includes: When it is determined that the initial diffusion area has an overlapping area, the corresponding overlapping area is used as a fused diffusion area, and the corresponding initial diffusion area is used as an impact diffusion area, and the sub-pollen concentration data of the impact diffusion area is obtained as the impact concentration data; According to the difference set between the affected diffusion area and the corresponding overlapping area, an independent diffusion area is obtained, and the affected concentration data corresponding to the affected diffusion area is used as the independent pollen concentration data of the independent diffusion area; Acquire a plurality of regional pixel values ​​corresponding to the fused diffusion region, and the regional area corresponding to each of the regional pixel values; Calculating the ratio of the area of ​​each region corresponding to the pixel value of each region to obtain the area ratio of the pixel value of the region; A calculation is performed based on the area proportion and the corresponding pixel value of the region to obtain a fused pixel value corresponding to the fused diffusion region.

8. The method according to claim 7, characterized in that The receiving of the regional data of the display area, customizing and adjusting the initial diffusion area according to the regional data to obtain an allergen area corresponding to the display area, and generating warning information based on the real-time location information of the user terminal and the allergen area and sending it to the user terminal includes: receiving regional data of the display area, the regional data including temperature, rainfall and wind speed; The theoretical diffusion distance is adjusted according to the temperature, the rainfall and the wind speed to obtain an actual diffusion distance; Based on the actual diffusion distance, pixel points of the area contour of the planting area are moved to generate an allergen area corresponding to the display area; The real-time location information of the user terminal is obtained, and it is determined that the real-time location information is in the allergen area, and a warning message is generated and sent to the user terminal.

9. A system for predicting and managing multi-dimensional pollen information, characterized in that: include: A training module is used to collect historical project data of the display area as an initial training set, input the initial training set into multiple first-level models in the initial forecast model for training to obtain intermediate prediction results and combine them, obtain the intermediate training set and input it into the second-level model for training to generate a pollen forecast model; A calculation module, for receiving a plurality of allergy factors set by a user terminal, obtaining real-time data of the corresponding display area based on the allergy factors, inputting the real-time data into a pollen prediction model to obtain pollen concentration data, and inputting the pollen concentration data into a pollen health impact model for calculation to obtain a pollen impact logarithm; A generating module, for generating an initial diffusion area corresponding to each planting area in the display area based on the allergen display strategy and the planting area of ​​each allergen in the display area when it is determined that the pollen influence logarithm is greater than a preset influence value; an adjustment module, configured to receive the regional data of the display area, customize and adjust the initial diffusion area according to the regional data, obtain the allergen area corresponding to the display area, and generate warning information based on the real-time location information of the user terminal and the allergen area and send it to the user terminal; When it is determined that the pollen impact logarithm is greater than a preset impact value, based on the allergen display strategy and the planting areas of each allergen in the display area, generating an initial diffusion area corresponding to each planting area in the display area, including: When it is determined that the pollen influence logarithm is greater than the preset influence value, the planting areas of each allergic factor in the display area and the sub-planting areas corresponding to each planting area are obtained; The total planting area corresponding to the display area is obtained, and the planting proportion corresponding to each of the allergic factors is obtained according to the ratio of the sub-planting area to the total planting area; Based on the product of the planting proportion and the corresponding pollen concentration data, sub-pollen concentration data corresponding to each of the planting areas is obtained; According to the ratio of the sub-pollen concentration data and the preset area concentration data, a concentration diffusion coefficient is obtained, and based on the product of the concentration diffusion coefficient and the preset diffusion distance, a theoretical diffusion distance corresponding to the planting area is obtained; The pixel points of the area outline of the planting area are moved based on the theoretical diffusion distance to generate initial diffusion areas corresponding to each planting area in the display area.

Images