Fusion traceability algorithm based on water environment multi-source monitoring data

Abstract

The invention discloses a fusion traceability algorithm based on water environment multi-source monitoring data, which comprises the steps of collecting multi-source data, collecting water environment data from a plurality of monitoring sources including a ground monitoring station, a remote sensing satellite, unmanned aerial vehicle patrol and a ground sensor network, and analyzing water quality spatial-temporal characteristics. Analyzing a change trend and a periodic rule of the water quality monitoring data under a long-time sequence; and environment influence analysis: based on water quality monitoring data, rainfall data and hydro meteorological data of the pollution time period, analyzing correlation between pollutant concentration and rainfall, flow and temperature through a Pearson's correlation coefficient and rank phase relation calculation method, and obtaining point source pollution research and judgment, non-point source research and judgment and comprehensive research and judgment conclusions. Through the multi-source data acquisition and progressive analysis method, the range of the pollution source can be quickly narrowed, timely and accurate decision support is provided for the environmental protection department, and the efficiency and accuracy of tracing the pollution source are greatly improved.

Description

Technical Field

[0001] The present invention relates to the fields of water quality monitoring, geographic information, data fusion, pollutant diffusion, pollution source identification, and environmental data analysis, and particularly relates to a fusion and traceability algorithm based on multi-source monitoring data of water environment. Background Art

[0002] Water environment monitoring is of great significance for water resource protection, water quality management, and pollution prevention and control. Traditional water environment monitoring methods mainly rely on data collection from a single source, such as water quality sampling analysis, on-site observation, etc. However, with the development of sensor technology and remote sensing technology, modern water environment monitoring systems can collect more abundant and diverse data types, including but not limited to satellite images, UAV aerial images, ground monitoring station data, mobile monitoring device data, etc. In order to extract valuable information from these multi-source data, researchers have developed a series of data fusion technologies, such as rule-based methods, statistical methods, machine learning methods, etc. These methods aim to improve the accuracy and reliability of monitoring data.

[0003] After a water pollution event occurs, traceability analysis technology is used to determine the source of pollutants, which is crucial for taking effective control measures. Existing traceability analysis methods mainly include model-based methods, isotope analysis methods, etc. Although certain progress has been made in the existing technology, there are still problems such as data consistency problems, uneven data quality, limited traceability accuracy, and insufficient data processing capabilities.

[0004] In summary, although the existing technology has achieved certain results in the field of water environment monitoring, there are still certain limitations in data fusion and traceability analysis. Therefore, developing a new fusion and traceability algorithm to solve the above problems is of great significance for improving the accuracy and feasibility of water pollution traceability. Summary of the Invention

[0005] The purpose of the present invention is to provide a fusion and traceability algorithm based on multi-source monitoring data of water environment, and solve the problem of limited traceability accuracy caused by data consistency, data quality and other problems in the existing technology. By integrating multi-source monitoring data and applying advanced data processing technology, without increasing the hardware cost, the accuracy and reliability of water environment monitoring are significantly improved, the accurate tracing of pollution sources is realized, and strong support is provided for water resource protection and environmental management, thereby solving the foregoing problems existing in the existing technology.

[0006] In order to achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0007] A fusion and traceability algorithm based on multi-source monitoring data of water environment, comprising the following steps:

[0008] S100, Multi-source data collection: Collect water environment data from multiple monitoring sources such as ground monitoring stations, remote sensing satellites, drone inspections, and ground sensor networks. The data includes water quality parameters, pollution source monitoring data, hydrological information, and meteorological data. Establish the time relationship, spatial relationship, and the relationship of "source-network-plant-outlet" for different data elements to provide a data basis for subsequent analysis;

[0009] S200, Analysis of water quality spatio-temporal characteristics: When the cross-section monitoring data exceeds the standard, analyze the change trend and periodic law of the water quality monitoring data in the long time series; Based on the pollution transmission algorithm model, analyze whether there is pollution contribution in the upstream and downstream cross-sections. If so, focus on analyzing the pollution emissions in the upstream river section; Use the method of sliding window + Granger causality test to infer the main river section where pollution occurs;

[0010] S300, Environmental impact analysis: Based on the water quality monitoring data, rainfall data, and hydrometeorological data during the pollution period, analyze the correlation between pollutant concentration and rainfall, flow, and temperature through the Pearson correlation coefficient and rank correlation coefficient calculation methods; If it is a strong correlation, it is determined that the pollution is caused by environmental factors, otherwise the environmental impact is excluded; At the same time, use time series data analysis to quantify the relationship between different variables and deeply understand the spatio-temporal variation law of water quality parameters;

[0011] S400, Point source pollution judgment: Based on the buffer analysis and the pollution transmission relationship of "source-network-plant-outlet-monitoring section", as well as the monitoring relationship between pollution sources, sewage treatment plants, and river outfalls, analyze whether there is a list of suspected enterprises with illegal emissions during the pollution period; Use a water quality fingerprint tracing instrument to analyze the water quality fingerprint at the pollution moment and dynamically compare it with the wastewater discharge fingerprint of relevant enterprises to recommend suspected enterprises with illegal emissions; Combine the above two methods to obtain a list of suspected enterprises with illegal emissions; Use the QUAL2K water quality model to simulate the migration and transformation process of pollutants in the water body and accurately lock the possible pollution sources;

[0012] S600, Non-point source judgment: Analyze the rainfall data, land use type, and topographic and geomorphic factors during the pollution period through a geographic information system, and combine historical data and empirical models to identify possible non-point source pollution areas; Use water quality monitoring data and remote sensing data to analyze the water quality change characteristics of non-point source pollution areas and the law of pollutant migration and diffusion; Combine on-site investigations and monitoring data, and use the QUAL2K water quality model, cluster analysis, and deep learning to comprehensively evaluate the impact degree of non-point source pollution on water quality and determine the source and main contribution area of non-point source pollution;

[0013] S700. Comprehensive Judgment Conclusion: Based on the results of the above spatio-temporal feature analysis, environmental impact analysis, point source pollution judgment, and non-point source judgment, a pollution conclusion is drawn. If there are upstream and downstream pollution contributions, pollution concentration changes caused by environmental impacts, discovery of illegal point source emissions, or discovery of suspected non-point source pollution areas, corresponding pollution tracing conclusions are formed. If no relevant judgment conclusions are found, the conclusion of no pollution found is recorded. Whether a conclusion is obtained or not, confirmation and analysis are supplemented by means of drones, unmanned ships, satellite remote sensing, and estuary monitoring technologies. The comprehensive judgment conclusion includes the conclusions of steps S100 to S600 and gives suggestions for tracing and investigation.

[0014] In some specific embodiments, in step S100, the water quality parameters include dissolved oxygen, pH value, and turbidity.

[0015] The pollution source monitoring data includes wastewater flow rate and pollutant concentration, and the hydrological information includes flow rate, flow velocity, and water level.

[0016] The meteorological data includes rainfall.

[0017] In some specific embodiments, in step S200, the pollution transmission algorithm model is constructed based on the river network topological structure and the principle of water flow dynamics, and is used to simulate the transmission process of pollutants in the river network.

[0018] In some specific embodiments, in step S400, the buffer zone analysis is specifically as follows: taking the monitoring section as the center, a buffer zone of a certain range is determined according to factors such as river flow velocity and pollutant diffusion coefficient, and the pollution transmission relationship between the pollution sources in the buffer zone and the monitoring section is analyzed.

[0019] In some specific embodiments, in step S600, deep learning uses a convolutional neural network to process remote sensing images, identify the land use types and vegetation coverage characteristics in the non-point source pollution area, and assist in evaluating non-point source pollution.

[0020] The beneficial effects of the present invention are as follows: The present invention discloses a fusion tracing algorithm based on multi-source monitoring data of water environment, including multi-source data collection, spatio-temporal feature analysis of water quality, environmental impact analysis, point source pollution judgment, non-point source judgment, and comprehensive judgment conclusion. The present invention has the following effects:

[0021] (1) Improvement in tracing efficiency and accuracy: Through multi-source data collection and progressive analysis methods, the scope of pollution sources can be quickly narrowed down, providing timely and accurate decision-making support for environmental protection departments, and greatly improving the efficiency and accuracy of pollution source tracing.

[0022] (2) Comprehensive and real-time water quality monitoring: The integrated use of multi-source data can more comprehensively reflect the water quality status, timely discover potential pollution problems, and enhance the comprehensiveness and real-time nature of water quality monitoring.

[0023] (3) Promote intelligent development: Provide strong technical support for water environment monitoring and management, and promote the intelligent development of water resource protection and environmental protection.

[0024] (4) Enhance emergency response capabilities: With the help of rapid and accurate pollution source tracing, environmental management departments can quickly take response measures, reduce the impact of pollution on the environment and human health, ensure ecological safety, and enhance emergency response capabilities.

[0025] (5) Reduce monitoring costs: Compared with traditional methods, the present invention adopts automated and intelligent means, reduces the input of manpower and material resources, improves the monitoring efficiency, and thus reduces the overall cost of environmental monitoring.

[0026] (6) Facilitate scientific research: The integration of multi-source monitoring data and the application of traceability algorithms provide rich and accurate data resources for water environment scientific research, and contribute to promoting scientific research progress in related fields. Brief Description of the Drawings

[0027] Figure 1 is a flowchart of a fusion traceability algorithm based on multi-source monitoring data of water environment of the present invention;

[0028] Figure 2 is a schematic diagram of the steps of the fusion traceability algorithm of the present invention. Detailed Embodiments

[0029] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0030] Refer to Figure 1 and Figure 2 A fusion traceability algorithm based on multi-source monitoring data of water environment shown, comprising the following steps:

[0031] S100, Multi-source data collection: Collect water environment data from multiple monitoring sources such as ground monitoring stations, remote sensing satellites, drone inspections, and ground sensor networks. The data includes water quality parameters (such as dissolved oxygen, pH value, turbidity, etc.), pollution source monitoring data (such as wastewater flow, pollutant concentration, etc.), hydrological information (such as flow rate, velocity, water level, etc.), and meteorological data (such as rainfall, etc.); establish the time relationship, spatial relationship, and the relationship of "source-network-plant-outlet" of different data elements to provide a data basis for subsequent analysis. In actual operation, utilize the existing water environment monitoring network, through data docking with meteorological departments, hydrological monitoring stations, satellite data receiving agencies, and various pollution source monitoring devices, to obtain meteorological, hydrological, remote sensing, water quality monitoring, and pollution source monitoring data. Adopt data cleaning and preprocessing technologies to unify the data time format, measurement unit, etc., and establish the time relationship, spatial relationship, and the relationship of "source-network-plant-outlet" among the data to ensure the consistency and availability of the data.

[0032] S200, Analysis of water quality spatio-temporal characteristics: When the cross-section monitoring data exceeds the standard, analyze the change trend and periodic law (such as diurnal characteristics, continuity, intermittency, bimodal characteristics, etc.) of the water quality monitoring data in the long time series; based on the pollution transmission algorithm model, analyze whether there is pollution contribution in the upstream and downstream cross-sections. If so, focus on analyzing the pollution discharge in the upstream river section; adopt the method of sliding window + Granger causality test to infer the main river section where pollution occurs. When the water quality monitoring data of a certain cross-section exceeds the standard, start the tracing program. Set a sliding window of appropriate size according to the actual situation, such as taking 1 hour as the window unit, and analyze the water quality data in multiple consecutive windows. Use the Granger causality test algorithm to analyze the causal relationship between the water quality data of each cross-section, and judge whether the pollutant is transmitted from upstream to downstream or there are other transmission paths, so as to determine the main river section where pollution occurs.

[0033] S300, Environmental impact analysis: Based on the water quality monitoring data, rainfall data, and hydro-meteorological data during the pollution period, analyze the correlation between pollutant concentration and rainfall, flow rate, and temperature through the Pearson correlation coefficient and rank correlation coefficient calculation methods; if it is a strong correlation, it is determined that the pollution is caused by environmental factors, otherwise the environmental impact is excluded; at the same time, adopt time series data analysis to quantify the relationship between different variables and deeply understand the spatio-temporal variation law of water quality parameters.

[0034] In this embodiment, collect the cross-section water quality, meteorological, and hydrological data during the pollution period, use the time series data analysis method to draw the curve of water quality parameters changing with time, and observe its change trend and periodic law. By calculating the Pearson correlation coefficient, analyze the correlation between pollutant concentration and variables such as rainfall, flow rate, and temperature. If the correlation coefficient is greater than the set threshold, it is determined that the environmental factor has a significant impact on the pollution, otherwise its impact is excluded.

[0035] S400, Point source pollution judgment: Based on buffer zone analysis and the pollution transmission relationship of "source-network-plant-outfall-monitoring section", as well as the monitoring relationships among pollution sources, sewage treatment plants, and river outfalls, analyze the list of suspected enterprises of point sources with illegal discharges during the pollution period; use a water quality fingerprint tracing instrument to analyze the water quality fingerprint at the pollution moment and conduct dynamic comparison with the wastewater discharge fingerprints of relevant enterprises to recommend suspected enterprises with illegal discharges; combine the above two methods to obtain a list of suspected enterprises with illegal discharges; use the QUAL2K water quality model to simulate the migration and transformation process of pollutants in the water body and accurately lock the possible pollution sources. It should be noted that in this embodiment, ("source" represents pollution sources, "network" represents sewage pipelines, "plant" represents sewage treatment plants, and "outfall" represents river outfalls).

[0036] In this embodiment, according to the determined polluted river section, combined with the "source-network-plant-outfall" relationship and pollution source monitoring data, input the data into the QUAL2K water quality model. Set parameters such as the initial concentration of pollutants, discharge rate, water velocity, and flow rate of the water body in the model to simulate the migration and transformation process of pollutants in the water body. By comparing the simulation results with the actual monitoring data, determine the possible pollution source enterprises and generate a list of suspected enterprises with illegal discharges.

[0037] S600, Non-point source judgment: Through Geographic Information System (GIS), analyze rainfall data, land use types, and topographic and geomorphic factors during the pollution period, and combine historical data and empirical models to identify possible non-point source pollution areas; use water quality monitoring data and remote sensing data to analyze the water quality change characteristics of non-point source pollution areas and the laws of pollutant migration and diffusion; combine on-site investigations and monitoring data, and use the QUAL2K water quality model, cluster analysis, and deep learning to comprehensively evaluate the impact degree of non-point source pollution on water quality and determine the sources and main contribution areas of non-point source pollution.

[0038] In this embodiment, use remote sensing satellites or drones to obtain high-resolution image data around the riverbank, use deep learning algorithms to process the images, and identify areas such as agricultural land and urban areas that may generate non-point source pollution. Combine the previous analysis results and meteorological data, and input the relevant data into the QUAL2K water quality model and cluster analysis algorithm. Through model simulation and cluster analysis, evaluate the impact degree of non-point source pollution on the polluted river section and determine the sources and main contribution areas of non-point source pollution.

[0039] S700. Comprehensive Judgment Conclusion: Based on the results of the above spatio-temporal feature analysis, environmental impact analysis, point source pollution judgment, and non-point source judgment, a pollution conclusion is drawn. If there are upstream and downstream pollution contributions, pollution concentration changes caused by environmental impacts, discovery of illegal point source emissions, or discovery of suspected non-point source pollution areas, corresponding pollution source tracing conclusions are formed. If no relevant judgment conclusions are found, it is recorded that no pollution conclusion is found. Whether a conclusion is obtained or not, drone, unmanned ship, satellite remote sensing, and estuary monitoring technical means are used for confirmation and analysis. The comprehensive judgment conclusion includes the conclusions of steps S100 to S600 and gives tracing and investigation suggestions.

[0040] Summarize the analysis results of the above steps, classify and organize the data according to pollution source type, location, scale, and possible reasons, etc., and generate a comprehensive tracing report. The report clearly points out point source pollution enterprises, non-point source pollution areas, and the impact of environmental factors on pollution, and gives targeted tracing and investigation suggestions according to the tracing results, such as suggesting strengthening supervision of specific enterprises and treating non-point source pollution in certain areas. At the same time, use technical means such as drones, unmanned ships, satellite remote sensing, and estuary monitoring to verify and supplement the analysis of the tracing results to ensure the accuracy of the conclusions.

[0041] In some specific embodiments, in step S100, the water quality parameters include dissolved oxygen, pH value, and turbidity.

[0042] The pollution source monitoring data includes wastewater flow rate and pollutant concentration, and the hydrological information includes flow rate, velocity, and water level.

[0043] The meteorological data includes rainfall.

[0044] In some specific embodiments, in step S200, the pollution transmission algorithm model is constructed based on the river network topological structure and the principles of water flow dynamics, and is used to simulate the transmission process of pollutants in the river network.

[0045] In some specific embodiments, in step S400, the buffer analysis is specifically as follows: taking the monitoring section as the center, a buffer zone of a certain range is determined according to factors such as river flow velocity and pollutant diffusion coefficient, and the pollution transmission relationship between the pollution sources in the buffer zone and the monitoring section is analyzed.

[0046] In some specific embodiments, in step S600, deep learning uses a convolutional neural network to process remote sensing images, identify the land use type and vegetation coverage characteristics in the non-point source pollution area, and assist in evaluating non-point source pollution.

[0047] The beneficial effects of the present invention are: The present invention discloses a fusion tracing algorithm based on multi-source monitoring data of water environment, including multi-source data collection, spatio-temporal feature analysis of water quality, environmental impact analysis, point source pollution judgment, non-point source judgment, and comprehensive judgment conclusion. The present invention has the following effects:

[0048] (1)Improved traceability efficiency and accuracy: Through multi-source data collection and progressive analysis methods, the scope of pollution sources can be quickly narrowed down, providing timely and accurate decision-making support for environmental protection departments, and greatly improving the efficiency and accuracy of pollution source tracing.

[0049] (2)Comprehensive and real-time water quality monitoring: The integrated use of multi-source data can more comprehensively reflect the water quality status, timely detect potential pollution problems, and enhance the comprehensiveness and real-time nature of water quality monitoring.

[0050] (3)Promote intelligent development: Provide strong technical support for water environment monitoring and management, and promote the intelligent development of water resource protection and environmental protection.

[0051] (4)Improve emergency response capabilities: With the help of rapid and accurate pollution source tracing, environmental management departments can quickly take response measures, reduce the impact of pollution on the environment and human health, ensure ecological safety, and improve emergency response capabilities.

[0052] (5)Reduce monitoring costs: Compared with traditional methods, the present invention adopts automated and intelligent means, reduces the input of manpower and material resources, improves the monitoring efficiency, and thus reduces the overall cost of environmental monitoring.

[0053] (6)Facilitate scientific research: The integration of multi-source monitoring data and the application of traceability algorithms provide rich and accurate data resources for water environment scientific research, and contribute to promoting the scientific research progress in related fields.

Claims

1. A fusion tracing algorithm based on multi-source monitoring data of water environment, characterized in that: The following steps are involved: S100, Multi-source data collection: Collect water environment data from multiple monitoring sources including ground monitoring stations, remote sensing satellites, drone inspections, and ground sensor networks. The data include water quality parameters, pollution source monitoring data, hydrological information, and meteorological data. Establish the temporal relationship, spatial relationship, and "source-network-factory-port" relationship of different data elements to provide a data basis for subsequent analysis. S200, water quality spatiotemporal characteristics analysis: when the section monitoring data exceeds the standard, analyze the change trend and periodicity of the water quality monitoring data in the long time series; based on the pollution transmission algorithm model, analyze whether there is pollution contribution in the upstream and downstream sections, and if so, focus on analyzing the pollution emission in the upstream river section; use the sliding window + Granger causality test method to infer the main river sections where pollution occurs; S300, Environmental Impact Analysis: Based on the water quality monitoring data, rainfall data and hydrological and meteorological data during the pollution period, the correlation between pollutant concentration and rainfall, flow and temperature is analyzed through the Pearson correlation coefficient and rank phase relationship calculation method; if there is a strong correlation, it is determined that the pollution is caused by environmental factors, otherwise the environmental impact is excluded; at the same time, time series data analysis is used to quantify the relationship between different variables and deeply understand the temporal and spatial variation of water quality parameters; S400, point source pollution assessment: Based on the buffer zone analysis and the pollution transmission relationship of "source-network-plant-outlet-monitoring section", as well as the monitoring relationship between pollution sources, sewage treatment plants, and sewage outlets into rivers, analyze whether there is a list of suspected enterprises with illegal discharge during the pollution period; use the water quality fingerprint tracer to analyze the water quality fingerprint at the time of pollution and dynamically compare it with the wastewater discharge fingerprint of related enterprises, and recommend suspected enterprises with illegal discharge; combine the above two methods to obtain a list of enterprises suspected of illegal discharge; use the QUAL2K water quality model to simulate the migration and transformation process of pollutants in the water body, and accurately identify possible pollution sources; S500, non-point source assessment: Analyze rainfall data, land use type, topographic factors during the pollution period through geographic information systems, and combine historical data and empirical models to identify possible non-point source pollution areas; use water quality monitoring data and remote sensing data to analyze the water quality change characteristics of non-point source pollution areas, as well as the laws of pollutant migration and diffusion; combine field surveys and monitoring data, use QUAL2K water quality model, cluster analysis and deep learning to comprehensively evaluate the impact of non-point source pollution on water quality, and determine the sources and main contributing areas of non-point source pollution; S600, comprehensive assessment conclusion: based on the results of the above-mentioned spatiotemporal feature analysis, environmental impact analysis, point source pollution assessment and non-point source assessment, a pollution conclusion is drawn; if there is upstream and downstream pollution contribution, environmental impact leads to changes in pollution concentration, illegal pollution point source emissions are found, and suspected non-point source pollution areas are found, then a corresponding pollution source tracing conclusion is formed; if no relevant assessment conclusion is found, then a record is made that no pollution is found; regardless of whether a conclusion is reached, it is confirmed and analyzed with the help of drones, unmanned ships, satellite remote sensing, and estuary monitoring technology; the comprehensive assessment conclusion includes the conclusions of steps S100 to S500, and provides suggestions for tracing the source.

2. The fusion tracing algorithm according to claim 1 is characterized in that: In step S100, the water quality parameters include dissolved oxygen, pH value, and turbidity; Pollution source monitoring data include wastewater flow and pollutant concentration, and hydrological information includes flow, velocity, and water level; Meteorological data includes rainfall.

3. The fusion tracing algorithm according to claim 1 is characterized in that: In step S200, the pollution transmission algorithm model is constructed based on the river network topology structure and water flow dynamics principles to simulate the transmission process of pollutants in the river network.

4. The fusion tracing algorithm according to claim 1 is characterized in that: In step S400, the buffer zone analysis is specifically as follows: taking the monitoring section as the center, determining a buffer zone within a certain range according to river flow velocity and pollutant diffusion coefficient factors, and analyzing the pollution transmission relationship between the pollution source in the buffer zone and the monitoring section.

5. The fusion tracing algorithm according to claim 1 is characterized in that: In step S600, the deep learning uses a convolutional neural network to process remote sensing images, identify land use types and vegetation coverage characteristics in the non-point source pollution area, and assist in assessing non-point source pollution.

Images