Environmental monitoring method based on simulation of atmospheric pollutant migration and transformation

Abstract

The present invention discloses an environmental monitoring method based on atmospheric pollutant migration and transformation simulation, which relates to the technical field of atmospheric pollutant monitoring, including: collecting and obtaining multiple particle characteristic information distributions; collecting and obtaining multiple meteorological data sequences, and constructing meteorological data spatial sequences; randomly selecting a first migrating particle, combining a first time node to obtain first meteorological data for screening to obtain first migration impact data; performing migration optimization on the first migrating particle to obtain a first migration parameter and perform position update to obtain a first updated position; obtaining the first updated position and the second meteorological data of the second time node, performing migration simulation on the first migrating particle, and completing the migration simulation on all particle sizes of particles, and generating particle environmental monitoring results in the target space. The present invention solves the technical problems of low monitoring accuracy and poor prediction accuracy in the prior art when dealing with complex and changeable atmospheric environments, and achieves the technical effect of improving monitoring accuracy and prediction accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of air pollutant monitoring, and in particular to an environmental monitoring method based on air pollutant migration and transformation simulation. Background Art

[0002] As global environmental problems become increasingly severe, especially the continued threat of air pollution to human health and the ecological environment, how to effectively monitor and manage the migration and diffusion of air pollutants has become a hot topic of global concern. Existing methods for monitoring air pollutants mainly rely on fixed sampling and monitoring equipment, which can only provide limited pollution data at specific locations and cannot fully capture the migration and diffusion of pollutants in the atmosphere. In addition, existing technologies usually lack real-time integration with meteorological conditions and cannot effectively reflect the impact of wind speed, wind direction, temperature, etc. on the migration of pollutants, resulting in the inability of monitoring data to accurately predict the diffusion trend of pollutants. Due to these technical limitations, existing methods have low monitoring accuracy and poor prediction accuracy when dealing with complex and changeable atmospheric environments, making it difficult to provide efficient support for pollution prevention and management. Summary of the invention

[0003] The present application provides an environmental monitoring method based on the simulation of the migration and transformation of atmospheric pollutants, which is used to solve the technical problems of low monitoring accuracy and poor prediction accuracy in the prior art when dealing with complex and changeable atmospheric environments.

[0004] In view of the above problems, the present application provides an environmental monitoring method based on the simulation of the migration and transformation of atmospheric pollutants.

[0005] The present application provides an environmental monitoring method based on atmospheric pollutant migration and transformation simulation, the method comprising:

[0006] Collect particle feature information at multiple locations in the target space to obtain multiple particle feature information distributions, wherein the particle feature information includes particle size and particle size ratio; collect meteorological data at the multiple locations in the target space within a preset time range in the future to obtain multiple meteorological data sequences, and construct a meteorological data space sequence of the target space; randomly select a particle of a first particle size in a first position as a first migration particle, obtain first meteorological data by indexing the meteorological data space sequence in combination with a first time node, perform screening, and obtain first migration influence data; perform migration optimization on the first migration particle according to the first migration influence data and the first particle size to obtain a first migration parameter, update the position of the first migration particle in the target space to obtain a first updated position, wherein optimization is performed in the migration optimization space according to the first migration influence data; obtain second meteorological data at the first updated position and a second time node by indexing in the meteorological data space sequence, continue to perform migration simulation on the first migration particle, and complete the migration simulation on particles of all particle sizes, obtain the final multiple simulated particle feature information distributions at the multiple locations, and generate the particle environment monitoring results in the target space.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The present application collects particle characteristic information at multiple positions in the target space and obtains multiple particle characteristic information distributions, wherein the particle characteristic information includes particle size and particle size ratio; collects meteorological data at multiple positions in the target space within a preset time range in the future, obtains multiple meteorological data sequences, and constructs a meteorological data space sequence of the target space; randomly selects a particle of a first particle size in a first position as a first migration particle, obtains the first meteorological data in combination with the first time node in the meteorological data space sequence index, performs screening, and obtains first migration influence data; performs migration optimization on the first migration particle according to the first migration influence data and the first particle size, obtains a first migration parameter, updates the position of the first migration particle in the target space, and obtains a first updated position, wherein optimization is performed in the migration optimization space according to the first migration influence data; obtains second meteorological data at the first updated position and the second time node by indexing in the meteorological data space sequence, continues to perform migration simulation on the first migration particle, and completes the migration simulation on particles of all particle sizes, obtains the final multiple simulated particle characteristic information distributions at multiple positions, and generates the particle environment monitoring results in the target space. The present invention solves the technical problems of low monitoring accuracy and poor prediction accuracy in the prior art when dealing with complex and changeable atmospheric environments. It collects particle feature information and meteorological data in the target space, optimizes and simulates the particle migration path in combination with the meteorological data spatial sequence, generates particle distribution information at multiple locations, provides accurate pollution diffusion trends and environmental monitoring results, and achieves the technical effect of improving monitoring accuracy and prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0010] Figure 1 A schematic diagram of a process flow of an environmental monitoring method based on atmospheric pollutant migration and transformation simulation provided in an embodiment of the present application;

[0011] Figure 2 A schematic diagram of a process for updating the position of a first migrating particle in an environmental monitoring method based on atmospheric pollutant migration and transformation simulation provided in an embodiment of the present application. DETAILED DESCRIPTION

[0012] The present application provides an environmental monitoring method based on the simulation of the migration and transformation of atmospheric pollutants, which is used to solve the technical problems of low monitoring accuracy and poor prediction accuracy in the existing technology when dealing with a complex and changeable atmospheric environment. By collecting the particle characteristic information and meteorological data in the target space, the particle migration path is optimized and simulated in combination with the spatial sequence of meteorological data, and the particle distribution information of multiple locations is generated, accurate pollution diffusion trends and environmental monitoring results are provided, thereby achieving the technical effect of improving monitoring accuracy and prediction accuracy.

[0013] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0014] It should be noted that any variations of the terms "include" and "have" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0015] Examples, such as Figure 1 As shown, the present application provides an environmental monitoring method based on atmospheric pollutant migration and transformation simulation, the method comprising:

[0016] Step S100: collecting particle feature information at multiple locations in the target space to obtain multiple particle feature information distributions, wherein the particle feature information includes particle size and particle size ratio.

[0017] In an embodiment of the present application, a sample of particulate matter in the air is detected by pre-deployed particulate matter detection equipment at multiple locations in the target space. One of the key features of the detection is the particle size of the particulate matter, that is, the size of the particles, expressed in microns (μm). For example, PM2.5 represents particulate matter with a particle size of less than 2.5 microns. In addition to the particle size, the equipment also detects the proportion of particles of different particle sizes, that is, the proportion of particles of each particle size in the overall particle matter. Through this process, multiple particle characteristic information distributions are obtained, wherein the multiple particle characteristic information distributions correspond one-to-one to multiple locations in the target space, and the particle characteristic information includes the particle size of the particle and the particle size ratio.

[0018] Furthermore, in the method provided in the embodiment of the application, the characteristic information of particles at multiple positions in the target space is collected to obtain the distribution of multiple particle characteristic information, and further includes:

[0019] At multiple locations in the target space, particle sampling detection is performed to obtain multiple particle size sets and the proportion of each particle size, and multiple particle feature information is obtained; based on the multiple particle sizes and particle size proportions in the multiple particle feature information, combined with the coordinates of the multiple locations, multiple particle feature information distributions are constructed.

[0020] In the embodiment of the present application, particle detection devices are deployed at multiple locations, and these devices are used to monitor the characteristic information of particles in the air in real time. The detection step involves measuring the particle size of the particles in the air, and calculating the proportion of each particle size based on the relative content of particles of different particle sizes. The particle size refers to the diameter of the particle, expressed in microns (μm), and the particle size proportion refers to the proportion of particles of a certain particle size in the total particle size.

[0021] At each detection point in the target space, the detection equipment will sample the surrounding air through sensors, identify and record particles of different sizes. By using technical equipment such as optical particle size analyzers and particle counters, the detection equipment can identify particles of different diameters and classify them to form multiple particle size sets. Each set contains particles of different sizes, such as PM2.5, PM10, etc. and their corresponding proportions. For example, at a certain detection point, PM2.5 particles accounted for 40%, while PM10 particles accounted for 30%, and the remaining 30% were larger or smaller particles. Through the above process, multiple particle feature information is obtained, where the particle size set at each position and the proportion of each particle size constitute the particle feature information at the corresponding position.

[0022] After obtaining the particle characteristic information of all locations, multiple particle characteristic information distribution models are constructed through spatial analysis tools or geographic information systems in combination with the spatial coordinates of each detection location, such as geographic longitude and latitude. This model shows the distribution of particle size and its proportion at multiple locations in the target space, that is, the distribution of multiple particle characteristic information.

[0023] Step S200: collecting meteorological data of the multiple locations in the target space within a preset time range in the future, obtaining multiple meteorological data sequences, and constructing a meteorological data space sequence of the target space.

[0024] In an embodiment of the present application, meteorological data of multiple locations in the target space within a preset time range in the future are obtained through meteorological monitoring data, including parameters such as wind direction, wind speed, temperature, humidity, and air pressure. Next, multiple meteorological data sequences are generated based on the meteorological forecast data of multiple locations. Each data sequence represents the meteorological changes of the location within a preset time range in the future. These sequences are composed of meteorological data at multiple time nodes, reflecting the changes in weather conditions at each location over a period of time in the future. For example, the wind speed and wind direction at a certain location may change significantly over time in the next few hours or days, and these changes directly affect the migration path and diffusion behavior of particulate matter. Then, multiple meteorological data sequences are combined with the geographic coordinates of each location to construct a spatial sequence of meteorological data in the target space. This spatial sequence shows the changing trends of meteorological conditions at each location in the target space at different time nodes.

[0025] Furthermore, in the method provided in the embodiment of the application, the meteorological data of the multiple locations in the target space within a preset time range in the future are collected to obtain multiple meteorological data sequences, and the meteorological data space sequence of the target space is constructed, which also includes:

[0026] Collect meteorological data of the multiple locations in the target space within a preset time range in the future to obtain multiple meteorological data sequences, wherein each meteorological data includes multiple wind direction information and wind force information; divide the multiple meteorological data sequences according to multiple time nodes to obtain multiple node meteorological data distributions, each node meteorological data distribution includes meteorological data of the multiple locations at a time node; construct multiple meteorological data spaces based on the multiple node meteorological data distributions to obtain the meteorological data space sequence of the target space.

[0027] In an embodiment of the present application, first, the meteorological data in the target space is extracted from multiple meteorological monitoring stations through an existing meteorological monitoring system. These monitoring stations are distributed in different locations in the target space, and continuously monitor and record meteorological parameters, such as wind direction, wind speed, temperature, humidity and air pressure. The monitoring data covers a preset time range in the future, such as meteorological monitoring data for the next few hours or days. By extracting these data from a meteorological database, a set of meteorological data related to time is generated. Next, the meteorological monitoring data of each location is organized in chronological order to form a meteorological data sequence. Each meteorological data sequence corresponds to a specific location, and records the meteorological information of the location at each time node within the preset time range in the future. Wherein, each meteorological data includes multiple wind direction information and wind force information.

[0028] Next, multiple meteorological data sequences are divided according to multiple time nodes. According to the pre-set time nodes within the future preset time range, the meteorological data of all monitoring locations are divided into multiple time node data sets. These sets represent the meteorological status of each monitoring location in the target space at different time points, forming multiple node meteorological data distributions. Each node meteorological data distribution contains data such as wind direction and wind speed at multiple locations.

[0029] Then, the meteorological data space is constructed by associating the meteorological data distribution of each node with the spatial coordinates of the monitoring location using geographic information system tools. This meteorological data space shows the distribution of meteorological conditions in the target space at a specific time node. For example, a node may show the spatial changes of wind direction and wind speed in the target space. Finally, these meteorological data spaces generated by time nodes are combined in chronological order to construct a meteorological data space sequence of the target space. This meteorological data space sequence shows the dynamic changes of meteorological conditions in the target space within a preset time range in the future.

[0030] Step S300: randomly selecting a particle of the first particle size in the first position as the first migration particle, obtaining first meteorological data in combination with the first time node in the meteorological data spatial sequence index, and screening to obtain first migration impact data.

[0031] In an embodiment of the present application, firstly, a certain position in the target space is randomly selected as the first position, which is the starting point of the migration of the particle. At this position, a particle of a specific particle size is randomly selected from a plurality of particle sizes, which is defined as the first particle size particle. For example, PM2.5 is selected as the first migrating particle. Next, in combination with the time nodes in the meteorological data spatial sequence, the meteorological data corresponding to the first time node is indexed. The first time node is the initial time node within the future preset time range, and there are N time nodes within the future preset time range, and the time difference between each time node is the same, for example, the time difference between the first time node and the second time node is 10 minutes. Wherein N is a positive integer greater than or equal to 3.

[0032] Then, based on the information such as wind direction and wind speed extracted from the meteorological data, these data are screened to generate the first migration impact data. The main purpose of the screening is to filter out the data that has little impact on the migration of particulate matter to improve the efficiency and accuracy of the migration simulation. For example, a migration impact threshold is set to delete meteorological data whose wind speed or wind direction is not sufficient to affect the migration of particulate matter, and retain the key meteorological information that has a significant impact on the migration of particulate matter. These screened data are the first migration impact data.

[0033] Furthermore, in the method provided in the embodiment of the application, a particle of the first particle size in the first position is randomly selected as the first migration particle, and the first meteorological data is obtained in combination with the first time node in the meteorological data spatial sequence index, and the first migration impact data is obtained by screening, and the method further includes:

[0034] Randomly select a first particle size particle of a first particle size in a first position within the distribution of the multiple particle characteristic information as a first migration particle; obtain the first meteorological data by indexing in the meteorological data spatial sequence according to the first position and the first time node; screen the multiple first wind direction information and the multiple first wind force information, as well as the gravity information in the first meteorological data according to the first particle size, obtain and delete the migration influence data that is less than a preset migration influence threshold, retain the other migration influence data, and obtain the first migration influence data.

[0035] In an embodiment of the present application, first, a specific position is randomly selected as the first position in the distribution of multiple particle characteristic information in the target space, and a particle of a specific particle size is selected from the particle information at the position, which is called the first particle size particle. For example, PM10 is randomly selected as the first migrating particle. Next, according to the selected first position and the specific first time node, an index operation is performed in the pre-constructed meteorological data space sequence. By indexing the sequence, the first meteorological data at the first time node is obtained.

[0036] Subsequently, the first meteorological data extracted from the meteorological data spatial sequence is screened in combination with the characteristics of the first particle size particles. The focus of the screening is to eliminate data with less impact on migration based on wind direction, wind speed and other meteorological conditions that affect the migration of particle matter. In this process, a preset migration impact threshold is set, that is, these data will only be retained when the wind direction, wind speed or gravity information has a sufficiently large impact on the migration of particle matter. For example, for larger particle sizes, stronger winds are required to cause significant migration, so data with wind speeds below a certain threshold will be deleted, leaving only migration impact data that has a significant impact on particle migration.

[0037] After this screening process, the meteorological information that is finally retained is the first migration impact data. These data cover the meteorological conditions that have a significant impact on the migration of particulate matter, such as wind direction, wind speed and gravity information at a specific time point.

[0038] Furthermore, in the method provided in the embodiment of the application, according to the first particle size, multiple first wind direction information and multiple first wind force information in the first meteorological data, as well as gravity information are screened, migration impact data less than a preset migration impact threshold is obtained and deleted, and other migration impact data is retained, and the method also includes:

[0039] According to the migration impact test data of particulate matter with different particle sizes, a sample particle size set, multiple sample wind force information sets, multiple sample wind force test data sets and multiple sample gravity test data sets are collected, wherein each sample wind force test data includes a migration vector under the sample wind force information; according to the multiple sample wind force test data sets and the sample gravity test data sets, multiple wind force test accuracy information and multiple gravity test accuracy information are calculated and obtained; the sample particle size set and the multiple sample wind force information sets are used as index inputs, and the multiple wind force test accuracy information and multiple gravity test accuracy information are used as index outputs to construct a migration impact accuracy classifier; the first particle size is combined with the multiple first wind force information input into the migration impact accuracy classifier for indexing, and multiple first wind force accuracy information and first gravity accuracy information are obtained; the wind force information or gravity information corresponding to the accuracy information less than the preset migration impact threshold is screened and deleted, and other migration impact data is retained to obtain the first migration impact data.

[0040] In an embodiment of the present application, a migration impact test is first performed based on particles of different particle sizes. In order to simulate and analyze the migration behavior of particles of different particle sizes under the action of wind and gravity, multiple sample data are collected. These sample data include a sample particle size set, that is, a representative sample of particles of multiple different particle sizes, such as PM2.5, PM10, etc.; and also include a sample wind information set, which records the wind speed and wind direction information under different wind conditions. At the same time, a sample wind test data set and a sample gravity test data set are collected. Each sample wind test data includes the migration behavior of particles under specific wind conditions, and the migration direction and distance are described by a migration vector. The gravity test data reflects the settling velocity and trajectory of particles under different gravity conditions.

[0041] Next, based on multiple sample wind test data sets and sample gravity test data sets, multiple wind test accuracy information and multiple gravity test accuracy information are calculated. Specifically, in the sample data, the migration vectors under each sample wind and gravity conditions are analyzed. Through statistical analysis, the vector with the largest proportion of migration vectors under each condition is determined, and it is considered that this is the most correct migration vector under the condition. For example, under a certain wind condition, if most particles migrate in a certain direction, the migration vector in that direction will be regarded as the most accurate migration vector. Then the wind test accuracy information is calculated. The wind test accuracy information is obtained by calculating the ratio of the number of occurrences of the most correct migration vector, that is, the migration vector with the largest proportion, to the total number of all migration vectors under the wind condition. This ratio indicates the extent to which the migration behavior of the particles conforms to the expected or main migration direction under the wind condition. The same logic applies to the gravity test accuracy information. According to the sample data under different gravity conditions, the migration vector with the largest proportion is statistically analyzed, and the gravity test accuracy is calculated by the ratio of the number of occurrences of the correct migration vector to the total number of all migration vectors.

[0042] In order to further optimize the accuracy of migration simulation, a migration impact accuracy classifier is constructed using these sample data. Specifically, the sample particle size set and the sample wind force information set are used as the input index of the classifier, and the accuracy information of the wind force test and the accuracy information of the gravity test are used as the output index to construct the migration impact accuracy classifier. The function of this classifier is to predict the accuracy of the corresponding wind force and gravity on the migration of particulate matter based on the input particle size and wind force conditions. Specifically, the sample particle size is first standardized, and the numerical range of different particle sizes is normalized to ensure the consistency of the scale of the input data. Important features of the wind force information, such as wind speed and wind direction, are extracted and converted into numerical forms that the classifier can process. The data after these feature processing will be used as the input index of the classifier. Next, a suitable machine learning algorithm is selected to build the classifier, such as a support vector machine. Then the classifier is trained using the collected input and output data. The processed sample particle size and wind force information are used as input indexes, and the wind force test accuracy information and gravity test accuracy information are used as output indexes. The training process uses supervised learning methods to optimize the internal parameters of the classifier through input-output pairs, so that it can accurately predict the impact of particle migration based on the input conditions. Through training, a migration impact accuracy classifier is obtained.

[0043] Then, in practical application, the first particle size particles and the first wind force information related thereto are input into the migration impact accuracy classifier for prediction and indexing. Through the analysis of the classifier, the first wind force accuracy information and the first gravity accuracy information are obtained.

[0044] Finally, data is screened based on the multiple first wind accuracy information and first gravity accuracy information obtained. If the wind or gravity accuracy information is less than the preset migration impact threshold, it means that these conditions have a weak impact on particle migration, and the wind or gravity data corresponding to these information are deleted. Those data that have a significant impact on migration are retained, and finally the optimized first migration impact data is obtained.

[0045] Step S400: performing migration optimization on the first migrating particles according to the first migration influencing data and the first particle size to obtain first migration parameters, and updating the positions of the first migrating particles in the target space to obtain first updated positions, wherein optimization is performed according to the first migration influencing data in the migration optimization space.

[0046] In an embodiment of the present application, when optimizing the migration of the first migrating particles according to the first migration influencing data and the first particle size, particle swarm optimization is used to optimize the migration of particles in the migration optimization space. The migration optimization space is an abstract space constructed based on factors such as the meteorological conditions of the target space, the physical properties of the particles, and environmental resistance, indicating that the particles can move at different positions in the space. In this space, the optimization algorithm gradually adjusts the migration parameters of the particles, such as the migration direction and speed, based on multiple migration influencing data to find the best migration path. During the optimization process, multiple migration parameter combinations are first initialized, and these parameter combinations are evaluated based on the characteristics of the first particle size and the first migration influencing data. The evaluation criterion is the migration fitness of the particles under given conditions, that is, whether these parameters can enable the particles to migrate effectively. Through multiple iterations, the optimization algorithm will continuously adjust these parameters until the migration parameter combination with the highest fitness is found.

[0047] Finally, the first migration parameters are determined after optimization, including the first migration direction and the first migration distance. These migration parameters will be used to update the position of the first migration particle so that the particles migrate according to the optimized path. Through these steps, the migration path of the particles is optimized, reflecting its optimal movement direction and position in the atmosphere. The first updated position is calculated based on the optimized migration parameters, that is, according to the new position of the first migration particle in the target space.

[0048] Further, such as Figure 2 As shown, in the method provided in the embodiment of the application, according to the first migration influence data and the first particle size, the first migrating particle is optimized for migration, the first migration parameter is obtained, and the position of the first migrating particle is updated in the target space, and further includes:

[0049] Acquire the migration vector space of particulate matter and construct a migration optimization space; initialize multiple optimized particles and multiple initial migration vectors in the migration optimization space; perform migration rationality analysis based on the multiple initial migration vectors in combination with the first particle size and the first migration influence data to obtain multiple initial migration fitnesses; adjust and update the initial migration vectors of the multiple optimized particles to obtain multiple updated migration vectors, and analyze and obtain multiple updated migration fitnesses; continue to optimize the migration vector of the first migrating particle in the migration optimization space until convergence, and obtain the migration vector with the maximum migration fitness as the first migration parameter, the first migration parameter including the first migration direction and the first migration distance; update the position of the first migrating particle based on the first migration parameter and the first position to obtain the first updated position.

[0050] In the embodiment of the present application, migration vector data are first obtained from historical experimental data. These vector data are obtained under different experimental conditions, including different particle sizes, different wind speeds and directions, and particle migration behaviors under the influence of gravity. The migration vector under each experimental condition represents the migration path, direction, and distance of the particles over a period of time. Based on these experimental results, a migration vector space is constructed, which contains the possible migration paths of particles under different environmental conditions.

[0051] Next, we construct the migration optimization space. The migration optimization space is based on the migration vector space and adds the search space of the optimization algorithm. This space allows the algorithm to search for the optimal migration path. To optimize in this space, we first initialize multiple optimization particles, each of which represents a potential migration path. At the same time, we initialize multiple initial migration vectors, which represent the initial position and movement direction of the optimization particles in the search space.

[0052] Based on these initial migration vectors, combined with the characteristics of the first particle size (such as particle size and physical properties) and the first migration impact data (such as wind and gravity data), these vectors are analyzed for migration rationality. The goal of the migration rationality analysis is to evaluate the rationality of these initial vectors under the current conditions, that is, whether these migration paths conform to the actual migration behavior of particulate matter. Through analysis, the initial migration fitness of each optimized particle is calculated. The fitness represents the quality of the migration path. The higher the fitness, the more the path conforms to the laws of physics.

[0053] After evaluating the initial fitness, the optimization phase begins. The initial migration vector of each optimized particle is adjusted and updated. Through an optimization algorithm, such as the particle swarm optimization algorithm, the position and movement direction of each particle are continuously adjusted according to its fitness to generate an updated migration vector. For example, the initial migration vector is randomly adjusted in direction length within the migration optimization space, and does not exceed the range of the migration optimization space. After each update, the updated migration fitness of each particle is recalculated to determine the rationality and quality of the new migration vector.

[0054] Next, continue to optimize the migration vector of the first migrating particle in the migration optimization space. Through multiple iterations, the algorithm gradually adjusts the migration vector of the particle until the fitness converges, that is, the optimal migration vector is found. This means that the algorithm has found the best migration path with the maximum fitness. The migration vector of this optimal path is called the first migration parameter, which includes the first migration direction and the first migration distance.

[0055] Finally, according to the obtained first migration parameter, the position of the first migrating particle is updated, and the first updated position is calculated.

[0056] Furthermore, in the method provided in the embodiment of the application, a migration rationality analysis is performed based on the multiple initial migration vectors in combination with the first particle size and the first migration influence data to obtain multiple initial migration fitnesses, and further includes:

[0057] According to the migration test data of particles of the first particle size in the historical time, a sample migration influence data set and a sample migration vector set are collected, and according to the accuracy information of each sample migration vector appearing under the sample migration influence data, a sample migration fitness set is labeled; the sample migration influence data set, the sample migration vector set and the sample migration fitness set are used to train a first migration rationality analyzer; the first migration rationality analyzer is used to perform migration rationality analysis on the multiple initial migration vectors in combination with the first migration influence data, respectively, to obtain multiple initial migration fitnesses.

[0058] In the embodiment of the present application, firstly, based on the migration test data of the first particle size collected in the historical time, a sample migration influence data set and a sample migration vector set are constructed. These test data are obtained under different environmental conditions and include the migration path and influencing factors of the particles, such as wind direction, wind speed, gravity, etc. The migration vector under each test condition represents the direction and distance of movement of the particles in the atmosphere, and the sample migration influence data reflects the main external conditions (such as wind and gravity) that affect the migration of particles.

[0059] Next, the accuracy information of each sample migration vector under its corresponding migration impact data is calculated. This process is performed by analyzing the vector with the largest proportion of migration vectors as the most accurate vector. Similar to the above calculation method, after obtaining the accuracy information of each sample, the sample is labeled according to this information to generate a sample migration fitness set, that is, the accuracy information is used as the migration fitness of the sample.

[0060] Then, the sample migration influence data set, the sample migration vector set and the sample migration fitness set are used to train the first migration rationality analyzer. Specifically, the sample migration influence data set, the sample migration vector set and the sample migration fitness set are used as inputs, and a machine learning algorithm such as a support vector machine is used for training to construct a first migration rationality analyzer. This analyzer predicts the rationality of the migration vector based on the input migration conditions. After the training is completed, multiple initial migration vectors are combined with the corresponding first migration influence data and input into the trained first migration rationality analyzer for analysis. The analyzer evaluates the rationality of these initial migration vectors based on the historical data patterns it has learned, and calculates the initial migration fitness of each initial vector to obtain multiple initial migration fitnesses.

[0061] Step S500: Index in the meteorological data spatial sequence to obtain the second meteorological data at the first update position and the second time node, continue to perform migration simulation on the first migrating particles, and complete the migration simulation on particles of all particle sizes to obtain the final distribution of multiple simulated particle characteristic information at the multiple positions, and generate the particle environment monitoring results in the target space.

[0062] In an embodiment of the present application, after performing a preliminary migration simulation and obtaining the first update position, the second meteorological data corresponding to the second time node is indexed from the meteorological data space sequence. The meteorological data space sequence is constructed by previous weather forecasts or meteorological monitoring data, and contains meteorological information at different locations in the target space at multiple time nodes, including wind speed, wind direction, temperature, humidity, etc. By indexing this sequence, the latest meteorological data at the first update position at the second time node is obtained.

[0063] Next, the migration simulation of the first migrating particle is continued using the new meteorological data. This simulation process is similar to the previous migration simulation, and the migration path and new updated position of the particles are calculated according to the current meteorological conditions. This process ensures that the migration path of the particles can be adjusted with the changes of time and environment under the changing meteorological conditions, so that the simulation results are closer to the real atmospheric dynamics. At the same time, the same migration simulation is performed on all particle sizes in the target space. Each particle size will be independently simulated for migration based on its unique physical properties and the meteorological conditions in which it is located. After multiple rounds of simulation, when the migration simulation in all time nodes is completed, the data of each time node are combined to generate multiple simulated particle feature information distributions at multiple locations in the target space. These information distributions show the characteristics of particles at different locations, such as the concentration and distribution range of particles of different particle sizes. These data reflect how particles migrate and diffuse in the atmosphere during the simulation time, as well as their final distribution state in the target space.

[0064] Finally, based on these simulation results, the particulate matter environmental monitoring results are generated. The monitoring results provide the diffusion of pollutants in the entire target space and show the concentration and distribution of particulate matter at different locations and time nodes.

[0065] Furthermore, in the method provided in the embodiment of the application, the second meteorological data of the first update position and the second time node is obtained by indexing in the meteorological data spatial sequence, the migration simulation of the first migrating particles is continued, and the migration simulation of particles of all particle sizes is completed, and the method further includes:

[0066] The second meteorological data of the first update position and the second time node are obtained by indexing in the meteorological data spatial sequence; the second meteorological data are screened to obtain second migration impact data, migration optimization and position update are performed, and multiple rounds of migration optimization are performed until the final time node of the future preset time range is reached to obtain the first final position; particles of the second particle size in the second position are randomly selected as second migration particles, multiple rounds of migration optimization and position update are performed, and migration simulation is completed for particles of all particle sizes in the multiple particle characteristic information distributions to obtain multiple simulated particle characteristic information distributions of the multiple positions.

[0067] In the embodiment of the present application, after the first migrating particle is preliminarily simulated for migration and the first updated position is obtained, the latest second meteorological data of the updated position at the second time node is indexed from the meteorological data spatial sequence. The meteorological data spatial sequence records the meteorological information at different time nodes and spatial positions, including wind direction, wind speed, temperature, humidity, etc.

[0068] Next, the extracted second meteorological data is screened to remove data that has little impact on particle migration. The screening criteria are based on the degree of impact of meteorological conditions on particle migration. For example, low wind speed may not significantly affect the movement of particles, so these data may be excluded. After screening, the retained data is called the second migration impact data, which will be used for subsequent migration optimization.

[0069] After having the second migration impact data, the first migration particle continues to be optimized and updated. This optimization process uses optimization algorithms, such as particle swarm optimization, to continuously adjust the migration vector to ensure that the migration path of the particles is the most reasonable under the current meteorological conditions. The optimization process may go through multiple rounds of iterations, each of which improves the position update of the particles by adjusting the migration vector until the preset accuracy requirements or the convergence of the simulation results are met. This process will continue until the final time node of the preset time range in the future is reached. At this time, the final position of the first migration particle will be determined as the first final position.

[0070] After completing the simulation of the first migrating particle, a particle of a different particle size in the second position is randomly selected as the second migrating particle. Similar to the processing method of the first migrating particle, the second migrating particle will undergo multiple rounds of migration optimization and position update. This process will continuously optimize its path until the final time node is reached. Once the simulation of particles of all particle sizes is completed, all simulation results are integrated, and the particle size of all particles in the target space is simulated to generate the distribution of particles at different time nodes and positions. Finally, the distribution of multiple simulated particle feature information at multiple positions is obtained.

[0071] In the embodiments of the present application, in summary, the embodiments of the present application have at least the following technical effects:

[0072] The present application collects particle characteristic information at multiple positions in the target space and obtains multiple particle characteristic information distributions, wherein the particle characteristic information includes particle size and particle size ratio; collects meteorological data at multiple positions in the target space within a preset time range in the future, obtains multiple meteorological data sequences, and constructs a meteorological data space sequence of the target space; randomly selects a particle of a first particle size in a first position as a first migration particle, obtains the first meteorological data in combination with the first time node in the meteorological data space sequence index, performs screening, and obtains first migration influence data; performs migration optimization on the first migration particle according to the first migration influence data and the first particle size, obtains a first migration parameter, updates the position of the first migration particle in the target space, and obtains a first updated position, wherein optimization is performed in the migration optimization space according to the first migration influence data; obtains second meteorological data at the first updated position and the second time node by indexing in the meteorological data space sequence, continues to perform migration simulation on the first migration particle, and completes the migration simulation on particles of all particle sizes, obtains the final multiple simulated particle characteristic information distributions at multiple positions, and generates the particle environment monitoring results in the target space. The present invention solves the technical problems of low monitoring accuracy and poor prediction accuracy in the prior art when dealing with complex and changeable atmospheric environments. It collects particle feature information and meteorological data in the target space, optimizes and simulates the particle migration path in combination with the meteorological data spatial sequence, generates particle distribution information at multiple locations, provides accurate pollution diffusion trends and environmental monitoring results, and achieves the technical effect of improving monitoring accuracy and prediction accuracy.

[0073] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0074] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0075] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. An environmental monitoring method based on simulation of atmospheric pollutant migration and transformation, characterized in that: The method comprises: Collecting particle feature information at multiple locations in the target space to obtain multiple particle feature information distributions, wherein the particle feature information includes particle size and particle size ratio; Collecting meteorological data of the plurality of locations in the target space within a preset future time range to obtain a plurality of meteorological data sequences, and constructing a meteorological data spatial sequence of the target space; Randomly select a particle of the first particle size in the first position as the first migration particle, obtain the first meteorological data by combining the first time node in the meteorological data spatial sequence index, and perform screening to obtain the first migration impact data; performing migration optimization on the first migrating particle according to the first migration influencing data and the first particle size to obtain a first migration parameter, and updating the position of the first migrating particle in the target space to obtain a first updated position, wherein optimization is performed according to the first migration influencing data in the migration optimization space; The second meteorological data at the first update position and the second time node are obtained by indexing in the meteorological data spatial sequence, the migration simulation of the first migrating particles is continued, and the migration simulation of particles of all particle sizes is completed to obtain the final distribution of characteristic information of multiple simulated particles at the multiple positions, and the particle environment monitoring result in the target space is generated; The method includes: obtaining the second meteorological data of the first update position and the second time node by indexing in the meteorological data spatial sequence, continuing to perform migration simulation on the first migrating particles, and completing the migration simulation on particles of all particle sizes, including: Obtain second meteorological data at the first update position and second time node by indexing in the meteorological data spatial sequence; The second meteorological data is screened to obtain second migration impact data, migration optimization and position update are performed, and multiple rounds of migration optimization are performed until the final time node of the future preset time range is reached to obtain the first final position; The second particle size particles in the second position are randomly selected as the second migration particles, and multiple rounds of migration optimization and position update are performed, and the migration simulation of all particle sizes within the multiple particle characteristic information distributions is completed to obtain the multiple simulated particle characteristic information distributions of the multiple positions.

2. The environmental monitoring method based on atmospheric pollutant migration and transformation simulation according to claim 1 is characterized in that: Collect particle feature information at multiple locations in the target space to obtain multiple particle feature information distributions, including: Performing particle sampling detection at multiple locations in the target space to obtain multiple particle size sets and the proportion of each particle size to obtain multiple particle feature information; Based on the multiple particle sizes and particle size ratios in the multiple particle feature information and in combination with the coordinates of the multiple positions, multiple particle feature information distributions are constructed.

3. The environmental monitoring method based on atmospheric pollutant migration and transformation simulation according to claim 1 is characterized in that: Collecting meteorological data of the multiple locations in the target space within a preset future time range to obtain multiple meteorological data sequences, and constructing a meteorological data space sequence of the target space, including: Collecting meteorological data of the multiple locations in the target space within a preset future time range to obtain multiple meteorological data sequences, wherein each meteorological data sequence includes multiple wind direction information and wind force information; Dividing the multiple meteorological data sequences according to multiple time nodes to obtain multiple node meteorological data distributions, each node meteorological data distribution includes meteorological data of the multiple locations at one time node; According to the distribution of meteorological data of the multiple nodes, multiple meteorological data spaces are constructed to obtain a meteorological data space sequence of the target space.

4. The environmental monitoring method based on atmospheric pollutant migration and transformation simulation according to claim 1 is characterized in that: Randomly select a particle of the first particle size in the first position as the first migration particle, obtain the first meteorological data in the meteorological data spatial sequence index in combination with the first time node, and perform screening to obtain the first migration impact data, including: Randomly selecting a first particle size particle of a first particle size in a first position from the plurality of particle characteristic information distributions as a first migrating particle; According to the first position and the first time node, indexing in the meteorological data spatial sequence to obtain first meteorological data; According to the first particle size, multiple first wind direction information and multiple first wind force information, as well as gravity information in the first meteorological data are screened to obtain and delete migration impact data that is less than a preset migration impact threshold, retain other migration impact data, and obtain first migration impact data.

5. The environmental monitoring method based on atmospheric pollutant migration and transformation simulation according to claim 4 is characterized in that: According to the first particle size, multiple pieces of first wind direction information and multiple pieces of first wind force information, as well as gravity information in the first meteorological data are screened to obtain and delete migration impact data that is less than a preset migration impact threshold, and retain other migration impact data, including: According to the migration impact test data of particles of different particle sizes, a sample particle size set, multiple sample wind force information sets, multiple sample wind force test data sets and multiple sample gravity test data sets are collected, wherein each sample wind force test data includes a migration vector under the sample wind force information; Calculate and obtain a plurality of wind test accuracy information and a plurality of gravity test accuracy information according to the plurality of sample wind test data sets and the sample gravity test data sets; Using the sample particle size set and multiple sample wind force information sets as index inputs, using the multiple wind force test accuracy information and multiple gravity test accuracy information as index outputs, to construct a migration impact accuracy classifier; Inputting the first particle size in combination with the plurality of first wind force information into the migration impact accuracy classifier for indexing, and obtaining a plurality of first wind force accuracy information and first gravity accuracy information; The wind information or gravity information corresponding to the accuracy information less than a preset migration impact threshold and other migration impact data are screened and deleted to obtain the first migration impact data.

6. The environmental monitoring method based on atmospheric pollutant migration and transformation simulation according to claim 1 is characterized in that: According to the first migration influence data and the first particle size, the first migrating particles are optimized for migration to obtain first migration parameters, and the positions of the first migrating particles are updated in the target space, including: Obtain the migration vector space of particles and construct the migration optimization space; Initializing a plurality of optimization particles and a plurality of initial migration vectors in the migration optimization space; Perform migration rationality analysis according to the multiple initial migration vectors in combination with the first particle size and the first migration influence data to obtain multiple initial migration fitnesses; Adjusting and updating the initial migration vectors of the plurality of optimized particles to obtain a plurality of updated migration vectors, and analyzing and obtaining a plurality of updated migration fitnesses; Continue to optimize the migration vector of the first migrating particle in the migration optimization space until convergence, and obtain the migration vector with the maximum migration fitness as the first migration parameter, wherein the first migration parameter includes a first migration direction and a first migration distance; The position of the first migrating particle is updated according to the first migration parameter and the first position to obtain a first updated position.

7. The environmental monitoring method based on atmospheric pollutant migration and transformation simulation according to claim 6 is characterized in that: A migration rationality analysis is performed according to the multiple initial migration vectors in combination with the first particle size and the first migration influence data to obtain multiple initial migration fitnesses, including: According to the migration test data of the particles of the first particle size in the historical time, a sample migration influence data set and a sample migration vector set are collected, and according to the accuracy information of each sample migration vector appearing under the sample migration influence data, a sample migration fitness set is obtained by annotation; Using the sample migration impact data set, the sample migration vector set and the sample migration fitness set to train a first migration rationality analyzer; The first migration rationality analyzer is used to perform migration rationality analysis on the multiple initial migration vectors in combination with the first migration impact data to obtain multiple initial migration fitnesses.

Images