Water pollution prevention and control tracing system based on big data

Abstract

The invention discloses a water pollution prevention and control tracing system based on big data, which comprises a cloud platform and a database, and is characterized in that the cloud platform is connected with the database; historical water area data including remote sensing data, layout data and water quality information are obtained through a cloud platform, the historical water area data are processed and integrated, the integrated historical water area data are stored in a database, and current water area data are collected. Big data analysis is carried out through current water area data and historical water area data in a database to obtain an influence area type and predicted water quality information, and emission source searching is carried out according to the influence area type and the current water area data to obtain a pollution emission source so as to realize prevention and tracing.

Description

Technical Field

[0001] The present invention belongs to the technical field of water pollution detection, and in particular relates to a water pollution prevention and control tracing system based on big data. Background Art

[0002] With the rapid development of industrialization, urbanization and agricultural modernization, water bodies are polluted by various organic and inorganic pollutants, including industrial wastewater, agricultural non-point source pollution, urban drainage systems, etc. These pollutants not only reduce the quality of water bodies, but may also have adverse effects on human health through drinking water, food chains and other pathways.

[0003] However, the amount of water data is large and messy, and is held in different databases or unit systems. The data is not interoperable and cannot be effectively and uniformly managed and used. At the same time, the prevention and control of water pollution can only be done through a comprehensive analysis of the water area, and professionals need to rely on manual experience and on-site testing to determine the emission source. It is impossible to conduct intelligent and rapid pollution source tracing at the data level. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes a water pollution prevention and control tracing system based on big data to solve the problems existing in the above-mentioned prior art.

[0005] To achieve the above objectives, the present invention provides a water pollution prevention and control tracing system based on big data, comprising:

[0006] A cloud platform and a database, wherein the cloud platform is connected to the database; wherein historical water area data is obtained through the cloud platform, wherein the water area data includes remote sensing data, layout data and water quality information, and the historical water area data is processed and integrated, and the integrated historical water area data is stored in the database, and current water area data is collected, and big data analysis is performed through the current water area data and the historical water area data in the database to obtain the type of affected area and predicted water quality information, and emission sources are searched according to the type of affected area and the current water area data to obtain the pollution emission sources for prevention and control and tracing.

[0007] Optionally, the water quality information includes organic matter concentration and different metal ion concentrations.

[0008] Optionally, in the cloud platform, the process of processing and integrating the water area data includes: completing the water area data, integrating the completed water area data, and integrating the water area data of the same type into the same folder.

[0009] Optionally, in the database, a data storage architecture is constructed based on the water area data, in which the root element is the region, and time is used as a sub-element under the region. The time sub-element is described in detail, wherein the detailed description includes the remote sensing data index corresponding to the time, the layout data index corresponding to the time, and the content of the environmental detection data.

[0010] Optionally, in the cloud platform, the process of obtaining the impact area type includes:

[0011] The remote sensing data in the current water area data and the remote sensing data in the historical water area data are segmented and extracted to obtain different types of regional ranges, and the correlation fitting is performed on the different types of regional ranges and water quality information. According to the fitting results, the impact area type is generated.

[0012] Optionally, in the cloud platform, based on water quality information in historical water area data and current water area data, predicted water quality information is obtained through a fitting method or a neural network model.

[0013] Optionally, in the cloud platform, the process of finding emission sources includes:

[0014] Obtain the initial emission substance concentrations of different emission sources according to the type of affected area, construct and initialize a population according to the initial emission substance concentration, wherein the population includes a number of individuals, wherein the individuals include the initial emission substance concentration; set an objective function, fitness and population update method, wherein the objective function is to minimize the square value of the difference between the simulated water quality information and the observed water quality information corresponding to different locations; according to the objective function, fitness and population update method, perform optimization through an optimization algorithm to generate the final emission substance concentration;

[0015] The final emission substance concentrations are sorted, and the pollution emission sources are obtained based on the sorting results.

[0016] Compared with the prior art, the present invention has the following advantages and technical effects:

[0017] Through the above technical solution, the present invention stores remote sensing data, layout data, and water quality data in an orderly manner through a database, unifies the above data into a database, and stores them through a data architecture. Big data analysis is performed on the data in the database to analyze the types of possible impact areas, and water pollution prevention and control is performed. On this basis, a water area model is constructed through real-time remote sensing data, and pollution emission sources are found through simulation of the water area model and optimization of emission sources. According to the pollution emission sources, the parks, living areas, or units that may cause water pollution are determined, and water pollution is effectively traced from the data level. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0019] Figure 1 Schematic diagram of the system structure of an embodiment of the present invention. DETAILED DESCRIPTION

[0020] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0021] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0022] like Figure 1 As shown, this embodiment provides a water pollution prevention and control tracing system based on big data, including:

[0023] A cloud platform and a database, wherein the cloud platform is used to integrate and process the water area data, and store the integrated and processed water area data in the database, and at the same time, perform relevant analysis on the water area data in the database to find the main impact area types and prediction information of polluted water quality, and prevent and control water pollution in the water area according to the main impact area types and prediction information, and after the polluted water quality information and prediction information are determined, perform relevant searches based on the facilities in the water area as virtual water pollution emission points, and perform simulation tracing to determine whether the corresponding facilities are pollution emission sources, so as to complete the tracing of pollution source emission points.

[0024] First, in the process of storing data in the database, the cloud platform places the water area data in an orderly manner into the database through relevant data collection, integration and processing. The water area data includes remote sensing data of the water area, industrial, commercial, agricultural and living layout data around the water area, environmental monitoring data, etc.; environmental monitoring data mainly includes water quality data, i.e. water quality information, such as organic matter concentration, concentration of different metal ions, etc. The above data can be obtained through existing acquisition methods such as downloading or collecting from the data published in existing databases or related websites, which will not be elaborated here.

[0025] The sources of the above data are different and numerous, and they need to be stored in the database through relevant processing and integration methods. First, the acquired remote sensing data, water area layout data, and environmental detection data are aligned in time scale, and after alignment, for the above three types of data at the same time, check whether there are missing values, and fill the missing values ​​through the missing remote sensing data filling method, mean interpolation method, or forward or backward interpolation method. After filling, the above data are integrated, and the same area is used as a directory. In different directories, different folders are constructed according to different data types, and specific data are stored in different folders. The above data is stored in the database in the above way, and it is convenient to extract relevant data. A relevant data storage architecture is constructed in the database, wherein the root element in the data storage architecture is the region, and the time is used as a sub-element under the region. The time sub-element is described in detail, wherein the detailed description includes the remote sensing data index corresponding to the time, the layout data index corresponding to the time, and the specific content of the environmental detection data. In the process of data extraction, there is no need to search for messy data. Through the above-mentioned orderly architecture or the above-mentioned integrated data storage method, the data can be extracted and searched in an orderly manner. When acquiring newly added data, the above content is stored according to the above storage method and architecture, and a corresponding historical database is generated to improve the orderliness of the data and avoid the disorderly storage of data from different sources.

[0026] The above database stores data in historical state. In order to facilitate the big data analysis of the current water area, the cloud platform needs to collect the corresponding data in the current state. Remote sensing data is obtained by taking photos of the water area through real-time remote sensing satellites. Environmental detection data is obtained through relevant real-time detection devices and sensors or detection sites, and transmitted to the cloud platform. The layout data around the water area can be identified through remote sensing data. After acquisition, it can also be stored in the database as newly added data.

[0027] In the process of big data analysis, the cloud platform first identifies the layout data of the water area around the current remote sensing data. During the identification process, the full convolutional neural network is used to extract the relevant regional segmentation of the current remote sensing data. The full convolutional neural network is used to extract the layout data of the water area around the remote sensing image, such as the pixel area of ​​the regional layout of industrial areas, commercial areas, agricultural areas and living areas. The convolution extraction is the low-order feature of the remote sensing image, and the image is processed according to the low-order features. In this process, the relevant structure of the full convolutional neural network is pre-set, and the relevant structure is the input layer, several convolutional layers and several deconvolutional layers in sequence. Finally, the final result is output in the output layer through the softmax function. According to the probability of different categories of layout data output by the softmax function, the category corresponding to the maximum probability is selected as the category that the pixel in the current image may correspond to, and the category is automatically annotated by different annotation methods to generate the corresponding layout data. The full convolutional neural network has been trained by the layout data and historical remote sensing data or other data sets in the database, and the training process is not repeated. After the layout data around the water area in the current remote sensing image is identified, the size of the corresponding area is counted, the layout data around the water area in the historical data is counted, and the size of the area is also counted. After the range of different categories of areas in the layout data is counted, the environmental detection data in the database, namely the historical water quality information and the current water quality information, is counted, and a correlation analysis is performed on different water quality information and different categories of area ranges. The range of different categories of areas is used as an independent variable, and the mean of different positions of a certain indicator in the water quality information is used as an independent variable to construct a multivariate regression model. The multivariate regression model is fit through the data in the database and the current data. After the fitting is completed, a relevant test is performed to obtain the main influencing area type of water quality information, and the main influencing area type is used as a prevention and control area.

[0028] At the same time, in the cloud platform, for water pollution corresponding to water quality information, the water quality information is set with a threshold or interval according to relevant standards. When the marked threshold or interval is exceeded, it is considered that water pollution occurs. If water pollution does not occur at present, the water quality information is predicted, and the water quality information in the future time is predicted by the regression analysis fitting curve of water quality information and time or the fitting relationship of a neural network model with current water quality information as input and water quality information as output for a period of time in the future, such as an LSTM network model (a network model trained by database data or other data). When water pollution occurs in the predicted water quality information, an early warning is given in time, and the above-mentioned main impact area types are prompted to prevent water pollution. In order to facilitate the visualization of the above-mentioned data by relevant personnel, the present invention also constructs a two-dimensional water area model according to the above-mentioned remote sensing data, different categories of areas and water quality information, and distinguishes them according to different colors according to the size of the water quality information. At the same time, different water quality information is displayed with a multi-layer visualization display scheme for visualization, and different water quality information is visualized to provide data guidance for the management of water areas.

[0029] At the same time, in order to more accurately determine the prevention and control location based on the geographical location in the cloud platform, the location that may cause water pollution is traced according to the location where the current water quality information is collected.

[0030] Among them, under the main impact area type, the emission site under this type is determined, and its emission substance concentration is used as the initial parameter to participate in the subsequent calculation. At the same time, the water quality information currently collected at different locations is used as the observation value to update the initial parameter to determine the emission site that causes water pollution. It should be noted that if the number of water quality information collection location points is less than a certain value (the location number threshold, which can be set according to manual experience), the correlation between the RGB or hyperspectral spectrum of the remote sensing data pixel points and the water quality information is constructed through the pixel points in the remote sensing data at the water quality information collection location. The correlation can be constructed according to a fitting curve or a deep learning model. Through the correlation between the water quality information and the remote sensing data pixel points, a point position in the water area distribution in the remote sensing data is randomly selected, and the water quality information of the position is fitted and generated. At the same time, the water quality information generated at the position is added to the observation value to search for the emission point that may cause water pollution.

[0031] In the above content in the cloud platform, the initial parameters are first determined. The initial parameters are the concentration of the emission substances at the emission sites of the main impact area types, including different metal ion concentrations and organic matter concentrations. The above initial parameters are simulated through the migration and diffusion components to simulate the simulated water quality information at the above different collection locations. In the process of generating simulated water quality information, according to the water area range in the remote sensing image data, the Visual modflow Flex software is used to build a two-dimensional water area model, and the relevant emission source locations and collection locations are set. The emission substance concentration is simulated at the emission source location, and the water quality information of the collection location during the simulation is calculated as the simulated water quality information through the relevant code components of the software water flow and pollutant migration. By comparing the simulated water quality information with the current water quality information, the initial parameters are updated according to the comparison results. After the update, the different collection simulated water quality information is recalculated. By repeating the above calculation process, the emission substance concentration at different emission sites is obtained, and the park or living area or park in the main impact area type corresponding to the emission site is checked to achieve water pollution tracing.

[0032] The concentration of emission substances from emission sources is calculated by improving the moth optimization algorithm. The specific steps of the optimization algorithm are as follows:

[0033] 1) Initialize the population, in which, in the process of optimizing the initial parameters or the emission concentration of the emission site, the emission concentration of different emission sites is used as the initial population, and the population contains different individuals or moth positions, and the individuals contain emission concentrations; its goal is to minimize the sum of the gaps between the simulated water quality information and the observed water quality information. The initialization individuals are based on artificial experience or random initialization data.

[0034] Set the relevant objective function, constraints and corresponding fitness, where the objective function is:

[0035] f(x)=min∑(A nm -A nm ') 2

[0036] Among them, A nm represents the mth observed water quality information at the nth position, A nm ' represents the mth simulated water quality information at the nth location calculated by the relevant code components of the simulation evaluation visualization software groundwater flow and pollutant transport. The constraints include that the water quality information at the location does not exceed its maximum value, and the fitness function is the inverse of the objective function.

[0037] 2) In the improved moth-to-fire algorithm, the moth needs to update and iterate the flames around it. The flames represent the local optimal solution, and the number of flames is determined. At the same time, during the iteration process, the number of flames needs to be adaptively reduced. The adaptive reduction mechanism formula is as follows:

[0038] FN = round(Nt*(N-1) / T)

[0039] Where: N is the initial number of flames; t is the current number of iterations; T is the maximum number of iterations, and FN represents the number of flames during the tth iteration.

[0040] 3) The fitness of different individuals is calculated according to the objective function, and the fitness is set to the inverse of the objective function: its fitness is g = 1 / f(x). The greater the fitness, the better the quality of the individual. During the calculation process, the data corresponding to the individual is added to the two-dimensional water model of the software as the concentration of emission substances at different emission locations. The flow and pollutant transport calculations are performed through the code components provided in the software to calculate the simulated water quality information at the collection location.

[0041] The fitness is sorted from large to small. After sorting, the corresponding number of flames is selected according to the sorting result, and the individuals in the fitness sorting are used as flames. The number of flames is calculated according to the above fitness reduction mechanism formula, and the current best flame is stored. The best flame is the individual with the highest number of flames in the fitness sorting.

[0042] 4) According to the above individual update method, or the moth position update method, the update is performed. In the update process, the initial position update method is first used for updating. In the initial position update method, the above moth position update method is adaptively updated through the inertia weight:

[0043] Mi=Di·e bθ ·cos(2πθ)+ω·Fj

[0044] Among them, b is a constant used to define the shape of the logarithmic spiral; θ is a random number between [-1,1]; Mi represents the i-th moth, that is, the individual; Fj represents the j-th flame; Di represents the distance between the i-th moth and the j-th flame, and the formula is the constant of the logarithmic spiral shape; the formula is: Di=|Fj-Mi|, ω is the inertia weight, which refers to the inertia weight idea of ​​the particle swarm algorithm. The larger the ω, the greater the search intensity in the global range of the algorithm, and the smaller the ω, the higher the search accuracy in the local range of the algorithm.

[0045]

[0046] Where t is the current iteration number; T is the maximum iteration number; a and d are constants respectively.

[0047] 5) After the update, update the moth position again:

[0048]

[0049] Among them, MI represents the label of the moth individual that is updated again, rand() represents a random number, and Levy represents the corresponding Levy random step.

[0050]

[0051] Among them, β is a constant, which has been set to 1.5, and the distribution number μ~N(0,σ 2 ), σ 2 represents the upper limit; random distribution number v~N(0,1);

[0052]

[0053] Where Г() represents the standard gamma function.

[0054] 6) Determine whether the flame data satisfies the requirement of only one flame remaining. If only one flame remains, directly output the final result;

[0055] 8) Determine whether the maximum number of iterations has been reached. If not, repeat the above steps 2-6. If reached, output the optimal moth individual, that is, the final emission substance concentration at different emission source locations.

[0056] Sort the emission substance concentrations of different emission source locations from large to small, select the emission source locations that exceed the emission standards or the first several emission source locations in the sorting as the pollution emission source locations, and check the parks, living areas, or units corresponding to the pollution emission source locations as the pollution source emission units. Check the parks, living areas, or units in the main impact area types corresponding to the emission locations to achieve water pollution traceability.

[0057] Through the above technical solution, the present invention stores remote sensing data, layout data, and water quality data in an orderly manner through a database, unifies the above data into a database, and stores them through a data architecture. Big data analysis is performed on the data in the database to analyze the types of possible impact areas, and water pollution prevention and control is performed. On this basis, a water area model is constructed through real-time remote sensing data, and pollution emission sources are found through simulation of the water area model and optimization of emission sources. According to the pollution emission sources, the parks, living areas, or units that may cause water pollution are determined, and water pollution is effectively traced from the data level.

[0058] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A water pollution prevention and control traceability system based on big data, It is characterized in that include: A cloud platform and a database, wherein the cloud platform is connected to the database; wherein historical water area data is obtained through the cloud platform, wherein the water area data includes remote sensing data, layout data and water quality information, and the historical water area data is processed and integrated, and the integrated historical water area data is stored in the database, and current water area data is collected, and big data analysis is performed through the current water area data and the historical water area data in the database to obtain the type of affected area and predicted water quality information, and emission sources are searched according to the type of affected area and the current water area data to obtain the pollution emission sources for prevention and control and tracing.

2. The system according to claim 1, It is characterized in that The water quality information includes organic matter concentration and different metal ion concentrations.

3. The system according to claim 1, It is characterized in that In the cloud platform, the process of processing and integrating the water area data includes: completing the water area data, integrating the completed water area data, and integrating the water area data of the same type into the same folder.

4. The system according to claim 1, It is characterized in that In the database, a data storage architecture is constructed based on the water area data. The root element in the data storage architecture is the region. Time is used as a sub-element under the region, and the time sub-element is described in detail, wherein the detailed description includes the remote sensing data index corresponding to the time, the layout data index corresponding to the time, and the content of the environmental detection data.

5. The system according to claim 1, It is characterized in that In the cloud platform, the process of obtaining the impact area type includes: The remote sensing data in the current water area data and the remote sensing data in the historical water area data are segmented and extracted to obtain different types of regional ranges, and the correlation fitting is performed on the different types of regional ranges and water quality information. According to the fitting results, the impact area type is generated.

6. The system according to claim 1, It is characterized in that In the cloud platform, based on the water quality information in the historical water area data and the current water area data, the predicted water quality information is obtained through a fitting method or a neural network model.

7. The system according to claim 1, It is characterized in that In the cloud platform, the process of finding emission sources includes: Obtain the initial emission substance concentrations of different emission sources according to the type of affected area, construct and initialize a population according to the initial emission substance concentration, wherein the population includes a number of individuals, wherein the individuals include the initial emission substance concentration; set an objective function, fitness and population update method, wherein the objective function is to minimize the square value of the difference between the simulated water quality information and the observed water quality information corresponding to different locations; according to the objective function, fitness and population update method, perform optimization through an optimization algorithm to generate the final emission substance concentration; The final emission substance concentrations are sorted, and the pollution emission sources are obtained based on the sorting results.

Images