An industrial park atmospheric environment pollution tracing method and system

Abstract

The application is suitable for the technical field of atmospheric environment protection, and provides an industrial park atmospheric environment pollution tracing method and system, which comprises the following steps: obtaining gas monitoring data of different spatial levels; determining the pollution gas concentration change trajectory of each position in the target industrial park according to the gas monitoring data of different spatial levels; and determining the pollution source position of the target industrial park based on the pollution gas concentration change trajectory of each position in the target industrial park. The pollution gas concentration change trajectory of each position in the target industrial park is determined through the gas monitoring data of different spatial levels, and the pollution source position of the target industrial park is determined by using the pollution gas concentration change trajectory of each position in the target industrial park, so that the problem that the atmospheric environment pollution source cannot be traced in the industrial park is solved.

Description

Technical Field

[0001] This invention belongs to the field of atmospheric environmental protection technology, and in particular relates to a method and system for tracing the source of atmospheric environmental pollution in industrial parks. Background Technology

[0002] With the development of urbanization and industrialization in my country, environmental damage has become increasingly serious. Driven by rapid socio-economic development, the number of urban industrial enterprises is constantly increasing, and their scale of development is also expanding. At the same time, the production and management of chemical industrial parks have a profound impact on the surrounding air quality, even causing health problems and inducing various diseases in nearby residents. Studies have shown that a normal person's daily respiratory volume is 10-12 cubic meters. Toxic and harmful chemicals in the atmosphere can enter the human body through the respiratory tract, causing physiological and sensory discomfort, and in severe cases, even poisoning and death.

[0003] Currently, there is no practical and effective technical method for tracing the sources of air pollution in industrial parks, and commonly used methods have many limitations. Therefore, there is an urgent need for a method to trace the sources of air pollution in industrial parks, solving the problem of not being able to trace the sources of air pollution in industrial parks. Summary of the Invention

[0004] This invention provides a method for tracing the source of atmospheric pollution in industrial parks, aiming to solve the problem of not being able to trace the source of atmospheric pollution within industrial parks. By using gas monitoring data at different spatial levels, the method determines the trajectory of pollutant gas concentration changes at various locations within the target industrial park. Using these trajectories, the location of the pollution source within the target industrial park is determined, thus solving the problem of not being able to trace the source of atmospheric pollution within industrial parks.

[0005] The present invention is implemented as follows: a method for tracing the source of atmospheric environmental pollution in industrial parks is provided, comprising the following steps:

[0006] Acquire gas monitoring data at different spatial levels;

[0007] Based on the gas monitoring data at different spatial levels, the trajectory of pollutant gas concentration changes at various locations within the target industrial park is determined.

[0008] Based on the trajectory of pollutant gas concentration changes at various locations within the target industrial park, the location of the pollution source in the target industrial park is determined.

[0009] Optionally, acquiring gas monitoring data at different spatial levels includes:

[0010] The first-space-level gas monitoring heat map sequence was obtained using satellite remote sensing methods;

[0011] Gas monitoring thermogram sequences at the second spatial level were obtained using low-altitude remote sensing methods;

[0012] A unique detection heatmap sequence at the third spatial level was obtained using a ground-based online sensor monitoring method.

[0013] Optionally, determining the trajectory of pollutant gas concentration at various locations within the target industrial park based on gas monitoring data from different spatial levels includes:

[0014] Determine the target area in the gas monitoring thermogram corresponding to the target industrial park at different spatial levels;

[0015] Based on the target area, a first target heatmap sequence is determined in the gas monitoring heatmap sequence of the first spatial level, a second target heatmap sequence is determined in the gas monitoring heatmap sequence of the second spatial level, and a third target heatmap sequence is determined in the gas monitoring heatmap sequence of the third spatial level.

[0016] Based on the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence, the trajectory of pollutant gas concentration at each location in the target industrial park is determined.

[0017] Optionally, determining the pollutant gas concentration trajectory at various locations within the target industrial park based on the first target heat map sequence, the second target heat map sequence, and the obtained third target heat map sequence includes:

[0018] Acquire real-time weather, building information, personnel activity, and vehicle activity data corresponding to the target industrial park;

[0019] Based on the time-series data, the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence, the trajectory of pollutant concentration at each location in the target industrial park is determined.

[0020] Optionally, determining the pollutant gas concentration trajectory at various locations within the target industrial park based on the time-series data, the first target heatmap sequence, the second target heatmap sequence, and the third target heatmap sequence includes:

[0021] The time-series data, the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence are input into the trajectory prediction model for trajectory prediction processing to obtain the pollutant gas concentration trajectory at each location in the target industrial park.

[0022] Optionally, before inputting the time-series data, the first target heatmap sequence, the second target heatmap sequence, and the third target heatmap sequence into the trajectory prediction model for trajectory prediction processing to obtain the pollutant gas concentration trajectory at each location in the target industrial park, the method further includes:

[0023] Acquire a sample dataset and a model to be trained. The sample data includes sample sequences and corresponding trajectory labels. The sample sequences include time-series data of the sample industrial park, a first sample heat map sequence, a second sample heat map sequence, and a third sample heat map sequence. The trajectory labels include the trajectory of pollutant gas concentration at various locations in the sample industrial park.

[0024] The model to be trained is subjected to supervised training using the sample dataset. Once training is complete, a trajectory prediction model is obtained.

[0025] Optionally, the training model includes a first feature extraction network, a second feature extraction network, a third feature extraction network, a fourth feature extraction network, a feature fusion network, and a linear regression network. The supervised training of the model to be trained using the sample dataset, upon completion of training, yields a trajectory prediction model, including:

[0026] The sample time series data, the first sample heatmap sequence, the second sample heatmap sequence, and the third sample heatmap sequence are respectively input into the first feature extraction network, the second feature extraction network, the third feature extraction network, and the fourth feature extraction network, and the first sample spatiotemporal features, the second sample spatiotemporal features, the third sample spatiotemporal features, and the fourth sample spatiotemporal features are respectively output.

[0027] The first sample spatiotemporal features, the second sample spatiotemporal features, the third sample spatiotemporal features, and the fourth sample spatiotemporal features are input into the feature fusion network, and the sample fusion features are output.

[0028] The sample fusion features are input into a linear regression network, and the trajectory results corresponding to the sample data are output.

[0029] The loss function between the trajectory result corresponding to the sample data and the trajectory label corresponding to the sample data is calculated. The optimization objective is to minimize the loss function. The parameters of the model to be trained are adjusted by the backpropagation algorithm. The parameter adjustment process is iterated. After the number of iterations reaches a preset number, the training is stopped, and the trajectory prediction model is obtained.

[0030] This invention also provides an air pollution source tracing device for industrial parks, comprising:

[0031] The acquisition module is used to acquire gas monitoring data at different spatial levels;

[0032] The first determining module is used to determine the trajectory of pollutant gas concentration changes at various locations in the target industrial park based on the gas monitoring data of different spatial levels.

[0033] The second determining module is used to determine the location of the pollution source in the target industrial park based on the trajectory of the change in pollutant gas concentration at various locations in the target industrial park.

[0034] This invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the computer program, it implements a method for tracing the source of atmospheric environmental pollution in an industrial park as described in any embodiment.

[0035] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements a method for tracing the source of atmospheric environmental pollution in an industrial park as described in any embodiment.

[0036] The beneficial effects achieved by this invention are as follows: This application obtains gas monitoring data at different spatial levels; based on the gas monitoring data at different spatial levels, it determines the trajectory of pollutant gas concentration changes at various locations within a target industrial park; and based on the trajectory of pollutant gas concentration changes at various locations within the target industrial park, it determines the location of the pollution source in the target industrial park. By determining the trajectory of pollutant gas concentration changes at various locations within a target industrial park using gas monitoring data at different spatial levels, and using this trajectory of pollutant gas concentration changes at various locations within the target industrial park to determine the location of the pollution source in the target industrial park, the problem of not being able to trace the source of atmospheric pollution in industrial parks is solved. Attached Figure Description

[0037] Figure 1 A flowchart of a method for tracing the source of atmospheric environmental pollution in an industrial park, provided as an embodiment of this application;

[0038] Figure 2 A schematic diagram of the structure of an atmospheric pollution source tracing device for industrial parks provided in this application embodiment;

[0039] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0041] Please refer to Figure 1 , Figure 1 A flowchart illustrating a method for tracing the source of atmospheric pollution in an industrial park, as provided in an embodiment of this application, is shown. The specific implementation principles of each step are as follows:

[0042] 101. Obtain gas monitoring data at different spatial levels.

[0043] In this embodiment of the invention, the aforementioned different spatial levels refer to different altitudes within the atmosphere. These spatial levels include upper atmosphere, lower and middle atmosphere, and near-ground atmosphere. Upper atmosphere can be understood as the space range with a vertical altitude of tens of kilometers or more; lower and middle atmosphere can be understood as the space range with a vertical altitude of several hundred to one thousand meters; and near-ground atmosphere can be understood as the space range with a vertical altitude of tens of meters or less.

[0044] The aforementioned gas monitoring data refers to the quantitative monitoring and recording of various gas components and their concentrations in the atmosphere. Gas monitoring data can also be used to study trends in atmospheric environmental change, predict weather, and assess air quality, which is of great significance to environmental protection and human health.

[0045] In this embodiment of the invention, the gas monitoring data can cover a variety of gases such as carbon dioxide, sulfur dioxide, nitrogen oxides, ozone, and particulate matter.

[0046] It should be noted that gas monitoring data at different spatial levels can be obtained by installing different types of gas monitoring equipment, such as ground meteorological stations, air quality monitoring stations, and satellite meteorological systems.

[0047] 102. Based on gas monitoring data at different spatial levels, determine the trajectory of pollutant gas concentration changes at various locations within the target industrial park.

[0048] In this embodiment of the invention, the gas monitoring data can cover a variety of gases such as carbon dioxide, sulfur dioxide, nitrogen oxides, ozone, and particulate matter.

[0049] The various locations within the aforementioned industrial parks can be various facilities, buildings, areas, or sites, such as factories, office buildings, warehouses, parking lots, etc.

[0050] The aforementioned pollutant concentration refers to the amount of pollutants contained in a unit volume or unit mass of air. It is usually expressed in units such as micrograms per cubic meter or milligrams per cubic meter. The higher the pollutant concentration, the more pollutants are contained per unit volume or unit mass of air, and the greater the harm to the environment and human health.

[0051] The above trajectory of pollutant gas concentration changes can be understood as the spatial concentration of pollutants changing over time.

[0052] It should be noted that by analyzing and processing gas monitoring data at different spatial levels, the changes in pollutant gas concentrations at different locations within the target industrial park can be determined, and a trajectory map of pollutant gas concentration changes can be drawn to better understand the sources, transmission, and transformation processes of air pollutants.

[0053] 103. Based on the trajectory of pollutant gas concentration changes at various locations within the target industrial park, determine the location of pollution sources within the target industrial park.

[0054] In this embodiment of the invention, each location in the industrial park can be a variety of facilities, buildings, areas or places, such as factories, office buildings, warehouses, parking lots, etc.

[0055] The above-mentioned trajectory of pollutant gas concentration changes can represent the spatial variation of pollutant gas concentration over time. Pollutant gas concentrations change with time, climate, and other factors.

[0056] In this embodiment of the invention, the location of the pollution source in the industrial park can be determined by the trajectory of the change in the concentration of pollutant gas in the factory, office building, warehouse, and parking lot.

[0057] It should be noted that by analyzing the trajectory of changes in pollutant gas concentration, we can understand the diffusion path of the pollutant gas and the location of the pollution source, and thus determine effective pollution control measures.

[0058] In this embodiment of the invention, gas monitoring data at different spatial levels are acquired; based on the gas monitoring data at different spatial levels, the trajectory of pollutant gas concentration changes at various locations within the target industrial park is determined; and based on the trajectory of pollutant gas concentration changes at various locations within the target industrial park, the location of the pollution source within the target industrial park is determined. By determining the trajectory of pollutant gas concentration changes at various locations within the target industrial park using gas monitoring data at different spatial levels, and by using the trajectory of pollutant gas concentration changes at various locations within the target industrial park, the location of the pollution source within the target industrial park is determined, thus solving the problem of not being able to trace atmospheric pollution sources within industrial parks.

[0059] Optionally, the steps for obtaining gas monitoring data at different spatial levels include: obtaining a gas monitoring heat map sequence for the first spatial level using satellite remote sensing; obtaining a gas monitoring heat map sequence for the second spatial level using low-altitude remote sensing; and obtaining a gas monitoring heat map sequence for the third spatial level using ground-based online sensor monitoring.

[0060] In this embodiment of the invention, the above-mentioned satellite remote sensing method is a method of using satellite sensors to acquire information about the Earth's surface or atmosphere, and then analyzing and processing this information to understand the condition and changing patterns of the Earth's surface and atmosphere.

[0061] The aforementioned first spatial level can be understood as the higher spatial level, typically referring to a spatial range at an altitude of tens of kilometers or more. Gas monitoring thermal map sequences for the higher spatial levels can be obtained using satellite remote sensing methods.

[0062] The aforementioned low-altitude remote sensing methods refer to methods of acquiring information about the Earth's surface or atmosphere by using remote sensing sensors on low-altitude platforms such as aircraft, drones, and balloons.

[0063] The aforementioned second spatial level can be understood as the low-to-mid-altitude space, typically referring to the spatial range between several hundred and one thousand meters in altitude. Gas monitoring thermal map sequences for the low-to-mid-altitude spatial level can be obtained using low-altitude remote sensing methods.

[0064] The aforementioned ground-based online sensor detection method refers to setting up online air pollution monitoring stations within industrial parks to monitor major pollutants in real time in order to understand their distribution and changing trends.

[0065] The aforementioned third spatial level can be understood as the space near the ground, typically referring to the area below tens of meters in altitude. Gas monitoring thermogram sequences for this near-ground spatial level can be obtained using ground-based online sensor monitoring methods.

[0066] The aforementioned gas monitoring heatmap is a type of graph used to visually display the distribution of gas concentrations.

[0067] It should be noted that the gas composition and concentration range monitored by gas monitoring heatmaps at different spatial levels are different, therefore, it is necessary to analyze and process the gas monitoring heatmaps at different spatial levels.

[0068] Optionally, the step of determining the trajectory of pollutant gas concentration at various locations within the target industrial park based on gas monitoring data at different spatial levels includes: determining the target area in the gas monitoring heatmap corresponding to the target industrial park at different spatial levels; based on the target area, determining a first target heatmap sequence in the gas monitoring heatmap sequence at the first spatial level, a second target heatmap sequence in the gas monitoring heatmap sequence at the second spatial level, and a third target heatmap sequence in the gas monitoring heatmap sequence at the third spatial level; and determining the trajectory of pollutant gas concentration at various locations within the target industrial park based on the first target heatmap sequence, the second target heatmap sequence, and the third target heatmap sequence.

[0069] In this embodiment of the invention, the aforementioned first spatial level can be a high-level spatial level, typically referring to a spatial range at an altitude of tens of kilometers or more. The aforementioned first target thermal map sequence refers to a gas monitoring thermal map sequence at a high-level spatial level.

[0070] The aforementioned second spatial level can be the low-to-mid altitude space, which usually refers to the space range between several hundred and one thousand meters in altitude. The aforementioned second target heat map sequence is indeed a gas monitoring heat map sequence of the low-to-mid altitude space level.

[0071] The aforementioned third spatial level can be the space near the ground, usually referring to the space range with a height of tens of meters or less. The aforementioned third target heat map sequence is essentially a gas monitoring heat map sequence of the space level near the ground.

[0072] The aforementioned target area can be understood as the area containing the gas monitoring heat map at the high-altitude space level, the gas monitoring heat map at the mid-to-low-altitude space level, and the gas monitoring heat map at the near-ground space level.

[0073] In this embodiment of the invention, the heat map sequence can be used to analyze the variation pattern of pollutant gas concentration at different locations within the target area.

[0074] It should be noted that by using gas monitoring heat map sequences at the upper-level space, the middle- and lower-level space, and the near-ground space, the concentration trajectories of pollutants at the upper-level space, the middle- and lower-level space, and the near-ground space can be determined.

[0075] Optionally, the step of determining the pollutant gas concentration trajectory at each location in the target industrial park based on the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence includes: acquiring real-time meteorological data, park buildings, personnel activities, and vehicle activities corresponding to the target industrial park; and determining the pollutant gas concentration trajectory at each location in the target industrial park based on the time-series data, the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence.

[0076] In this embodiment of the invention, the aforementioned real-time meteorology refers to the real-time monitoring and recording of meteorological data for a certain area, including real-time data of meteorological elements such as temperature, humidity, air pressure, wind speed, and wind direction.

[0077] The aforementioned buildings in the industrial park refer to buildings within the park, including various types of buildings such as factories, office buildings, warehouses, and laboratories.

[0078] The aforementioned personnel activities refer to personnel activities within the industrial park, including the activities of workers, managers, visitors, and other personnel within the industrial park.

[0079] The aforementioned vehicles refer to vehicle activities within the industrial park, including vehicles entering and exiting the park, logistics and delivery vehicles, and employee commuter vehicles. These vehicle activities generate exhaust emissions, such as carbon dioxide, carbon monoxide, nitrogen oxides, and volatile organic compounds, which can impact air quality. Therefore, industrial park management requires the management and control of vehicle activities, and the implementation of appropriate measures to reduce their impact on the atmospheric environment.

[0080] The aforementioned time series data refers to a data set arranged in chronological order. It is usually continuous and ordered time series data, which can be used to record and analyze changes in environmental factors.

[0081] The various locations within the aforementioned industrial parks can be various facilities, buildings, areas, or sites, such as factories, office buildings, warehouses, parking lots, etc.

[0082] It should be noted that the concentration trajectories of pollutants at various locations can be used to guide environmental governance and pollution prevention efforts. For example, it can help pinpoint the location of pollution sources and take targeted measures to control the emission and spread of pollutants, thereby protecting the environment and public health.

[0083] Optionally, the step of determining the pollutant gas concentration trajectory at each location in the target industrial park based on time-series data, the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence includes: inputting the time-series data, the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence into the trajectory prediction model for trajectory prediction processing to obtain the pollutant gas concentration trajectory at each location in the target industrial park.

[0084] In this embodiment of the invention, the trajectory prediction model refers to a model constructed based on historical trajectory data and other auxiliary information for predicting the future trajectory of a target object.

[0085] The aforementioned trajectory prediction can be understood as predicting the temporal and spatial trends of pollutant gas concentration. Historical monitoring data and historical trajectory data can be used to predict and simulate the dispersion of pollutants over a future period. Based on the prediction and simulation results, the changing trends and possible distribution of pollutant gas concentrations within the target area can be inferred.

[0086] The various locations within the aforementioned industrial parks can be various facilities, buildings, areas, or sites, such as factories, office buildings, warehouses, parking lots, etc.

[0087] It should be noted that the target heat map sequence and time series data are used as input, and the trajectory prediction model is used to predict the trajectory of pollutant gas concentration at different locations in the target industrial park.

[0088] Optionally, it also includes: acquiring a sample dataset and a model to be trained, wherein the sample dataset includes sample sequences and corresponding trajectory labels, the sample sequences include sample time-series data of the sample industrial park, a first sample heat map sequence, a second sample heat map sequence, and a third sample heat map sequence, and the trajectory labels include the pollutant gas concentration trajectories at various locations in the sample industrial park; and supervised training of the model to be trained using the sample dataset, and upon completion of training, a trajectory prediction model is obtained.

[0089] In this embodiment of the invention, the above-mentioned sample dataset can be understood as one of the data sets collected from big data.

[0090] The aforementioned "model to be trained" refers to a machine learning model that has not yet been trained. A model to be trained requires a supervised training process, during which the model learns and adjusts based on data from a sample dataset and is tested on a test set to evaluate its predictive accuracy.

[0091] The aforementioned pollutant concentration refers to the amount of pollutants contained in a unit volume or unit mass of air. It is usually expressed in units such as micrograms per cubic meter or milligrams per cubic meter. The higher the pollutant concentration, the more pollutants are contained per unit volume or unit mass of air, and the greater the harm to the environment and human health.

[0092] The above trajectory of pollutant gas concentration changes can be understood as the spatial concentration of pollutants changing over time.

[0093] Supervised training, as described above, refers to training a model using known training data, enabling the model to accurately predict new input data. Supervised training includes linear regression, logistic regression, decision trees, random forests, neural networks, and more.

[0094] It should be noted that in supervised training, the training data consists of input data and corresponding output data. The model learns the relationship between the input data and the output data to obtain a function that can map the input data to the correct output. During the training process, the model will continuously adjust its parameters based on the error of the training data to improve the accuracy of the prediction.

[0095] Optionally, the training model includes a first feature extraction network, a second feature extraction network, a third feature extraction network, a fourth feature extraction network, a feature fusion network, and a linear regression network. Supervised training is performed on the model to be trained using a sample dataset. The steps to obtain the trajectory prediction model after training include: inputting the sample time-series data, the first sample heatmap sequence, the second sample heatmap sequence, and the third sample heatmap sequence into the first feature extraction network, the second feature extraction network, the third feature extraction network, and the fourth feature extraction network, respectively, and outputting the first sample spatiotemporal features, the second sample spatiotemporal features, and the third sample spatiotemporal features, respectively. The first, second, third, and fourth sample spatiotemporal features are input into a feature fusion network, and the fused features are output. The fused features are input into a linear regression network, and the trajectory results corresponding to the sample data are output. The loss function between the trajectory results corresponding to the sample data and the trajectory labels corresponding to the sample data is calculated. The loss function is minimized as the optimization objective. The parameters of the model to be trained are adjusted through the backpropagation algorithm. The parameter adjustment process is iterated. After the number of iterations reaches a preset number, the training is stopped, and the trajectory prediction model is obtained.

[0096] In this embodiment of the invention, the aforementioned feature extraction network refers to a type of neural network used in deep learning to extract useful features from raw data. The feature extraction network can employ common neural networks such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), or Autoencoders.

[0097] The aforementioned feature fusion network refers to a type of neural network that fuses multiple feature maps into a new feature map.

[0098] The linear regression network described above is used to solve regression problems, which involve predicting a continuous output value based on input data. The basic structure of a linear regression network consists of one or more linear layers, each containing several neurons and corresponding weight parameters. The input to a linear regression network is a vector, and the output is a scalar.

[0099] The loss function mentioned above is a function used to measure the degree of difference between the model's predicted results and the actual results.

[0100] It's important to note that when training the trajectory prediction model, we need to input a sample dataset into the model for training, obtaining the model's predictions for that dataset. Then, we compare the trajectory labels corresponding to the sample dataset with the model's predictions, calculating the degree of difference between the two, i.e., the loss function. The smaller the loss function, the closer the model's predictions are to the actual results, and the higher the model's prediction accuracy.

[0101] The backpropagation algorithm described above is a supervised learning algorithm, often used to train multilayer perceptrons. The backpropagation algorithm mainly consists of two iterative steps (incentive propagation and weight update) until the network's response to the input reaches a predetermined target range.

[0102] In this embodiment of the invention, the backpropagation algorithm propagates the loss function from the output layer to the input layer, calculates the impact of each parameter on the loss function, and then updates the parameters according to the gradient descent method to minimize the loss function.

[0103] It should be noted that the method for tracing the source of atmospheric environmental pollution in industrial parks provided in this embodiment of the invention can be applied to devices such as smartphones, computers, and servers that can perform source tracing of atmospheric environmental pollution in industrial parks.

[0104] Please refer to Figure 2 , Figure 2 The diagram shown is a schematic diagram of an atmospheric pollution source tracing device for industrial parks provided in an embodiment of this application. The atmospheric pollution source tracing device for industrial parks includes:

[0105] The first acquisition module 201 is used to acquire gas monitoring data at different spatial levels;

[0106] The first determining module 202 is used to determine the trajectory of pollutant gas concentration change at various locations in the target industrial park based on the gas monitoring data of different spatial levels.

[0107] The second determining module 203 is used to determine the location of the pollution source in the target industrial park based on the trajectory of the change in the concentration of pollutant gas at various locations in the target industrial park.

[0108] Optionally, the acquisition module 201 includes:

[0109] The first acquisition submodule is used to acquire the gas monitoring heat map sequence at the first spatial level through satellite remote sensing methods;

[0110] The second acquisition submodule is used to acquire gas monitoring thermal map sequences at the second spatial level using low-altitude remote sensing methods.

[0111] The third acquisition submodule is used to acquire the unique detection heat map sequence of the third spatial level based on the ground-based online sensor monitoring method.

[0112] Optionally, the first determining module 202 includes:

[0113] The first determining submodule is used to determine the target area of ​​the target industrial park in the gas monitoring heat map corresponding to different spatial levels;

[0114] The second determining submodule is used to determine a first target heatmap sequence in the gas monitoring heatmap sequence of the first spatial level, a second target heatmap sequence in the gas monitoring heatmap sequence of the second spatial level, and a third target heatmap sequence in the gas monitoring heatmap sequence of the third spatial level, based on the target area.

[0115] The third determining submodule is used to determine the trajectory of pollutant gas concentration at various locations in the target industrial park based on the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence.

[0116] Optionally, the third determining submodule includes:

[0117] The acquisition unit is used to acquire real-time weather, building information, personnel activity, and vehicle activity data corresponding to the target industrial park.

[0118] The determining unit is used to determine the trajectory of pollutant concentration at each location in the target industrial park based on the time-series data, the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence.

[0119] Optionally, the determining unit is further configured to input the time-series data, the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence into the trajectory prediction model for trajectory prediction processing, so as to obtain the pollutant gas concentration trajectory at each location in the target industrial park.

[0120] Optionally, the device further includes:

[0121] The second acquisition module is used to acquire sample datasets and models to be trained. The sample data includes sample sequences and corresponding trajectory labels. The sample sequences include time-series data of sample industrial parks, a first sample heat map sequence, a second sample heat map sequence, and a third sample heat map sequence. The trajectory labels include pollutant gas concentration trajectories at various locations in the sample industrial parks.

[0122] The training module is used to perform supervised training on the model to be trained using the sample dataset. Once training is complete, a trajectory prediction model is obtained.

[0123] Optionally, the training module includes:

[0124] The first output submodule is used to input the sample time series data, the first sample heatmap sequence, the second sample heatmap sequence, and the third sample heatmap sequence into the first feature extraction network, the second feature extraction network, the third feature extraction network, and the fourth feature extraction network, respectively, and output the first sample spatiotemporal features, the second sample spatiotemporal features, the third sample spatiotemporal features, and the fourth sample spatiotemporal features, respectively.

[0125] The second output submodule is used to input the first sample spatiotemporal features, the second sample spatiotemporal features, the third sample spatiotemporal features, and the fourth sample spatiotemporal features into the feature fusion network, and output the sample fusion features.

[0126] The third output submodule is used to input the sample fusion features into the linear regression network and output the trajectory results corresponding to the sample data.

[0127] The processing submodule is used to calculate the loss function between the trajectory result corresponding to the sample data and the trajectory label corresponding to the sample data. With minimizing the loss function as the optimization objective, the parameters of the model to be trained are adjusted through the backpropagation algorithm. The parameter adjustment process is iterated. After the number of iterations reaches a preset number, the training is stopped, and the trajectory prediction model is obtained.

[0128] It should be noted that the method for tracing the source of atmospheric environmental pollution in industrial parks provided in this embodiment of the invention can be applied to computers, servers and other equipment that can perform source tracing of atmospheric environmental pollution in industrial parks.

[0129] Please refer to Figure 3 , Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 3 As shown, it includes: a memory 302, a processor 301, and a computer program for tracing the source of atmospheric environmental pollution in industrial parks, stored on the memory 302 and capable of running on the processor 301, wherein:

[0130] The processor 301 is used to call the computer program stored in the memory 302 and perform the following steps:

[0131] Acquire gas monitoring data at different spatial levels;

[0132] Based on the gas monitoring data at different spatial levels, the trajectory of pollutant gas concentration changes at various locations within the target industrial park is determined.

[0133] Based on the trajectory of pollutant gas concentration changes at various locations within the target industrial park, the location of the pollution source in the target industrial park is determined.

[0134] Optionally, the acquisition of gas monitoring data at different spatial levels performed by processor 301 includes:

[0135] The first-space-level gas monitoring heat map sequence was obtained using satellite remote sensing methods;

[0136] Gas monitoring thermogram sequences at the second spatial level were obtained using low-altitude remote sensing methods;

[0137] A unique detection heatmap sequence at the third spatial level was obtained using a ground-based online sensor monitoring method.

[0138] Optionally, the process executed by processor 301 to determine the trajectory of pollutant gas concentration at various locations within the target industrial park based on gas monitoring data at different spatial levels includes:

[0139] Determine the target area in the gas monitoring thermogram corresponding to the target industrial park at different spatial levels;

[0140] Based on the target area, a first target heatmap sequence is determined in the gas monitoring heatmap sequence of the first spatial level, a second target heatmap sequence is determined in the gas monitoring heatmap sequence of the second spatial level, and a third target heatmap sequence is determined in the gas monitoring heatmap sequence of the third spatial level.

[0141] Based on the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence, the trajectory of pollutant gas concentration at each location in the target industrial park is determined.

[0142] Optionally, the process executed by processor 301 to determine the pollutant gas concentration trajectory at various locations in the target industrial park based on the first target heat map sequence, the second target heat map sequence, and the obtained third target heat map sequence includes:

[0143] Acquire real-time weather, building information, personnel activity, and vehicle activity data corresponding to the target industrial park;

[0144] Based on the time-series data, the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence, the trajectory of pollutant concentration at each location in the target industrial park is determined.

[0145] Optionally, the processor 301 executes the step of determining the pollutant gas concentration trajectory at various locations in the target industrial park based on the time-series data, the first target heatmap sequence, the second target heatmap sequence, and the third target heatmap sequence, including:

[0146] The time-series data, the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence are input into the trajectory prediction model for trajectory prediction processing to obtain the pollutant gas concentration trajectory at each location in the target industrial park.

[0147] Optionally, before inputting the time-series data, the first target heatmap sequence, the second target heatmap sequence, and the third target heatmap sequence into the trajectory prediction model for trajectory prediction processing to obtain the pollutant gas concentration trajectory at each location in the target industrial park, the method executed by the processor 301 further includes:

[0148] Acquire a sample dataset and a model to be trained. The sample data includes sample sequences and corresponding trajectory labels. The sample sequences include time-series data of the sample industrial park, a first sample heat map sequence, a second sample heat map sequence, and a third sample heat map sequence. The trajectory labels include the trajectory of pollutant gas concentration at various locations in the sample industrial park.

[0149] The model to be trained is subjected to supervised training using the sample dataset. Once training is complete, a trajectory prediction model is obtained.

[0150] Optionally, the training model includes a first feature extraction network, a second feature extraction network, a third feature extraction network, a fourth feature extraction network, a feature fusion network, and a linear regression network. The processor 301 performs supervised training on the model to be trained using the sample dataset. Upon completion of training, a trajectory prediction model is obtained, including:

[0151] The sample time series data, the first sample heatmap sequence, the second sample heatmap sequence, and the third sample heatmap sequence are respectively input into the first feature extraction network, the second feature extraction network, the third feature extraction network, and the fourth feature extraction network, and the first sample spatiotemporal features, the second sample spatiotemporal features, the third sample spatiotemporal features, and the fourth sample spatiotemporal features are respectively output.

[0152] The first sample spatiotemporal features, the second sample spatiotemporal features, the third sample spatiotemporal features, and the fourth sample spatiotemporal features are input into the feature fusion network, and the sample fusion features are output.

[0153] The sample fusion features are input into a linear regression network, and the trajectory results corresponding to the sample data are output.

[0154] The loss function between the trajectory result corresponding to the sample data and the trajectory label corresponding to the sample data is calculated. The optimization objective is to minimize the loss function. The parameters of the model to be trained are adjusted by the backpropagation algorithm. The parameter adjustment process is iterated. After the number of iterations reaches a preset number, the training is stopped, and the trajectory prediction model is obtained.

[0155] It should be noted that the electronic devices provided in the embodiments of the present invention can be applied to devices such as smartphones, computers, and servers that can perform source tracing methods for atmospheric environmental pollution in industrial parks.

[0156] The electronic device provided in this embodiment of the invention can realize all the processes implemented in the above-described method embodiment for tracing the source of atmospheric environmental pollution in industrial parks, and can achieve the same beneficial effects. To avoid repetition, it will not be described again here.

[0157] This application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements various processes of the embodiment of the method for tracing the source of atmospheric environmental pollution in industrial parks provided by this invention, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0158] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for tracing the source of atmospheric environmental pollution in industrial parks, characterized in that, Includes the following steps: Acquire gas monitoring data at different spatial levels; Based on the gas monitoring data at different spatial levels, the trajectory of pollutant gas concentration changes at various locations within the target industrial park is determined. Based on the trajectory of pollutant gas concentration changes at various locations within the target industrial park, the location of the pollution source in the target industrial park is determined. The acquisition of gas monitoring data at different spatial levels includes: The first-space-level gas monitoring heat map sequence was obtained using satellite remote sensing methods; Gas monitoring thermogram sequences at the second spatial level were obtained using low-altitude remote sensing methods; The gas monitoring thermogram sequence of the third space level was obtained based on the ground-based online sensor detection method; The determination of the pollutant gas concentration trajectory at various locations within the target industrial park includes: Determine the target area in the gas monitoring thermogram corresponding to the target industrial park at different spatial levels; Based on the target area, a first target heatmap sequence is determined in the gas monitoring heatmap sequence of the first spatial level, a second target heatmap sequence is determined in the gas monitoring heatmap sequence of the second spatial level, and a third target heatmap sequence is determined in the gas detection heatmap sequence of the third spatial level. Based on the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence, the trajectory of pollutant gas concentration at each location in the target industrial park is determined; The step of determining the pollutant gas concentration trajectory at various locations within the target industrial park based on the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence includes: Acquire real-time weather, building information, personnel activity, and vehicle activity data corresponding to the target industrial park; Based on the time-series data, the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence, the trajectory of pollutant gas concentration at each location in the target industrial park is determined; The step of determining the pollutant gas concentration trajectory at various locations within the target industrial park based on the time-series data, the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence includes: The time-series data, the first target heat map sequence, the second target heat map sequence, and the third target heat map sequence are input into the trajectory prediction model for trajectory prediction processing to obtain the pollutant gas concentration trajectory at each location in the target industrial park.

2. The method for tracing the source of atmospheric environmental pollution in industrial parks as described in claim 1, characterized in that, Before inputting the time-series data, the first target heatmap sequence, the second target heatmap sequence, and the third target heatmap sequence into the trajectory prediction model for trajectory prediction processing to obtain the pollutant gas concentration trajectory at each location in the target industrial park, the method further includes: Obtain a sample dataset and a model to be trained. The sample dataset includes sample sequences and corresponding trajectory labels. The sample sequences include time-series data of the sample industrial park, a first sample heat map sequence, a second sample heat map sequence, and a third sample heat map sequence. The trajectory labels include the pollutant gas concentration trajectories at various locations in the sample industrial park. The model to be trained is subjected to supervised training using the sample dataset. Once training is complete, a trajectory prediction model is obtained.

3. The method for tracing the source of atmospheric environmental pollution in industrial parks as described in claim 2, characterized in that, The model to be trained includes a first feature extraction network, a second feature extraction network, a third feature extraction network, a fourth feature extraction network, a feature fusion network, and a linear regression network. Supervised training of the model to be trained using the sample dataset is performed. Upon completion of training, a trajectory prediction model is obtained, including: The sample time series data, the first sample heatmap sequence, the second sample heatmap sequence, and the third sample heatmap sequence are respectively input into the first feature extraction network, the second feature extraction network, the third feature extraction network, and the fourth feature extraction network, and the first sample spatiotemporal features, the second sample spatiotemporal features, the third sample spatiotemporal features, and the fourth sample spatiotemporal features are respectively output. The first sample spatiotemporal features, the second sample spatiotemporal features, the third sample spatiotemporal features, and the fourth sample spatiotemporal features are input into the feature fusion network, and the sample fusion features are output. The sample fusion features are input into a linear regression network, and the trajectory results corresponding to the sample data are output. The loss function between the trajectory result corresponding to the sample data and the trajectory label corresponding to the sample data is calculated. The optimal goal is to minimize the loss function. The parameters of the model to be trained are adjusted by the backpropagation algorithm. The parameter adjustment process is iterated. After the number of iterations reaches a preset number, the training is stopped, and the trajectory prediction model is obtained.

4. An air pollution source tracing device for industrial parks, used to implement the method as described in claim 1, characterized in that, include: The acquisition module is used to acquire gas monitoring data at different spatial levels; The first determining module is used to determine the trajectory of pollutant gas concentration changes at various locations in the target industrial park based on the gas monitoring data of different spatial levels. The second determining module is used to determine the location of the pollution source in the target industrial park based on the trajectory of the change in pollutant gas concentration at various locations in the target industrial park.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, When the processor executes the computer program, it implements a method for tracing the source of atmospheric environmental pollution in industrial parks as described in any one of claims 1 to 3.

6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the method for tracing the source of atmospheric environmental pollution in industrial parks as described in any one of claims 1 to 3.

Images