Water area ecological environment supervision system based on artificial intelligence

Abstract

The invention discloses a water area ecological environment supervision system based on artificial intelligence, and relates to the technical field of ecological environment supervision, and the system comprises a pollution monitoring module which is used for obtaining the geographic information of a water area, building a plurality of water quality monitoring points, and obtaining pollution data; the reverse deduction module is used for acquiring a hydrological model and meteorological data of a water area and surrounding enterprise data and determining a suspected pollution source; the searching and reporting module is used for dispatching an underwater robot and an unmanned aerial vehicle to search the suspected pollution source according to the suspected pollution source, generating a searching report and sending the searching report to a supervision department; the ecological acquisition module is used for respectively calling the underwater robot and the unmanned aerial vehicle to carry out normalized patrol in a water area, constructing a biodiversity database and constructing a hydrology and water quality database; and the ecological visualization module is used for constructing a water area twinborn map and generating a visual ecological environment supervision map. The method has the effect of improving the efficiency and real-time performance of monitoring the ecological environment of the water area.

Description

Technical Field

[0001] The present application relates to the technical field of water ecological environment supervision, and in particular to a water ecological environment supervision system based on artificial intelligence. Background Art

[0002] Water ecological environment supervision is to analyze and process various data such as water quality and hydrology of a water area in order to achieve management and optimization. In the existing technology, when collecting and managing water quality, hydrology and other data, the data of the relevant water area is generally obtained through manual sampling or reports from enthusiastic people, and then further inspection and processing are carried out based on the results obtained. In this way, when pollution occurs in the water area or changes in ecological biological species, it is difficult to detect these changes in a timely manner. After these changes occur, it takes a lot of time and energy to find the source of these changes or problems, which is very inefficient. Summary of the Invention

[0003] The purpose of the present invention is to provide an artificial intelligence-based water ecological environment monitoring system to solve the problems raised in the above background technology.

[0004] This application provides an artificial intelligence-based water ecological environment supervision system, which includes: Pollution monitoring module: used to obtain geographical information of the water area, establish multiple water quality monitoring points based on the geographical information, obtain water quality data, and obtain pollution data based on the water quality data; Inverse deduction module: used to obtain the hydrological model, meteorological data and surrounding enterprise data of the water area, and perform inverse deduction of the pollution path based on the hydrological model, the pollution data, the meteorological data and the surrounding enterprise data to determine the suspected pollution source; Investigation and reporting module: used to dispatch underwater robots and drones to investigate the suspected pollution sources, obtain investigation data, identify the pollution source enterprises based on the investigation data, generate an investigation report, and send the investigation report to the regulatory authorities; Ecological collection module: used to call underwater robots and drones to conduct regular patrols in the water area, and collect underwater biological data, shore biological data, and hydrological and water quality data during the patrol process, build a biodiversity database based on the underwater biological data and the shore biological data, and build a hydrological and water quality database based on the hydrological and water quality data; Ecological visualization module: used to construct a water area twin map based on the geographic information, add the biodiversity database and the hydrological and water quality database to the water area twin map, and generate a visual ecological environment supervision map.

[0005] Preferably, the steps of obtaining geographical information of a water area, establishing a plurality of water quality monitoring points based on the geographical information, obtaining water quality data, and obtaining pollution data based on the water quality data are specifically as follows: Obtaining geographic information, and obtaining the main water area, the number of tributaries, and the length of the tributaries according to the geographic information; Based on the scope of the main water area, a uniform monitoring network is constructed to patrol the main water area and identify high-risk areas that require key monitoring; Determining high-risk monitoring points based on the high-risk areas, and adjusting the uniform monitoring network based on the high-risk monitoring points to generate a monitoring network; Extracting the intersection positions of the network intersections of the monitoring network, and establishing water quality monitoring points within the main water area according to the intersection positions; According to the number and length of the tributaries, a water quality monitoring point is established on each tributary at a preset length interval; Each of the water quality monitoring points is called to monitor the water quality in the water area to obtain water quality data, and pollution data is obtained based on the water quality data.

[0006] Preferably, the steps of obtaining a hydrological model, meteorological data, and data of surrounding enterprises in a water area, and reversely deducing the pollution path based on the hydrological model, the pollution data, the meteorological data, and the data of surrounding enterprises to determine the suspected pollution source are specifically as follows: Obtaining a hydrological model, meteorological data, and surrounding enterprise data of the water area, and obtaining the direction of water flow in the water area based on the hydrological model and the meteorological data; Based on the data of the surrounding enterprises, the types of waste generated by the surrounding enterprises during their production processes are obtained, and the chemical composition of the emissions is obtained based on the types of waste; Extracting the pollution data to obtain target monitoring locations and pollution chemical data in the pollution data; Determine the approximate direction and range of the pollution source based on the water flow direction and the target monitoring location; Comparing the pollution chemical data with the emission chemical composition to obtain a comparison value, and screening the plurality of surrounding enterprises according to the comparison value to obtain a plurality of suspected target enterprises; Based on the general direction range and multiple suspected target enterprises, the pollution path is reversed to determine the suspected pollution source.

[0007] Preferably, after the step of reverse deducing the pollution path based on the approximate direction range and the plurality of suspected target enterprises to determine the suspected pollution source, the method further includes: If no suspected target enterprise is obtained after screening the plurality of surrounding enterprises according to the comparison value, the possible pollution source type is obtained according to the geographic information and the approximate direction range; According to the possible pollution source type, further obtaining the formation mode of the pollution source, and determining the scope of the pollution source according to the formation mode; Based on the geographic information and the scope of the pollution source, the pollution path is reversed, and a search is conducted within the scope of the pollution source to determine the suspected pollution source.

[0008] Preferably, the steps of dispatching underwater robots and drones to survey the suspected pollution sources, obtaining survey data, determining the pollution source enterprises based on the survey data, and generating a survey report are specifically as follows: constructing underwater inspection routes and aerial inspection routes based on the geographic information; dispatching an underwater robot to conduct an inspection according to the underwater inspection route to obtain underwater inspection data; Dispatching drones to conduct surveys according to the aerial inspection routes to obtain aerial inspection data; Obtaining a first spatiotemporal scale of the underwater robot and a second spatiotemporal scale of the unmanned aerial vehicle, and aligning the first spatiotemporal scale with the second spatiotemporal scale when the underwater robot and the unmanned aerial vehicle conduct a survey; Aligning the underwater survey data and the aerial survey data to generate an underwater-aerial linkage data table; Based on the underwater-air linkage data table, pollution source tracing is carried out to obtain the pollution source enterprises and generate a survey report.

[0009] Preferably, after the step of constructing the underwater inspection route and the aerial inspection route according to the geographic information, the method further includes: Based on the aerial inspection route, obtaining a route area passed by the aerial inspection route, and determining whether there is a sensitive area in the route area; If it is determined that there is a sensitive area in the route area, obtain the regulatory department of the sensitive area, apply for flight permission from the regulatory department, and fly after obtaining the flight permission; otherwise, change the inspection route to bypass the sensitive area; determining, based on the route area, whether there are any aerial obstacles in the route area; If it is determined that there is an aerial obstacle, the volume information, height information, and type information of the aerial obstacle are obtained; The aerial inspection route is adjusted according to the volume information, the altitude information, and the type information.

[0010] Preferably, the steps of respectively calling an underwater robot and a drone to conduct regular patrols in the water area and collecting underwater biological data, shore biological data, and hydrological and water quality data during the patrols are specifically as follows: Based on the uniform monitoring network, a main water patrol route for underwater robots and drones is constructed, a tributary water patrol route is generated according to the number of tributaries and the length of the tributaries, and a water patrol route is generated according to the main water patrol route and the tributary water patrol route; Binding an underwater robot and a drone to generate a patrol group, and releasing the patrol group according to a preset time interval to perform regular patrols along the water patrol route; Obtaining the edge outline of the water area based on the geographic information, and expanding outwards based on the edge outline to obtain the shore ecological range; Generate a shore patrol route based on the shore ecological range, and release a drone at preset time intervals to perform regular patrols along the shore patrol route; During patrols, underwater robots collect underwater images, underwater chemical composition, and underwater water flow information, while drones collect surface and shore images. Underwater biological data is obtained based on the underwater image and the surface image, shore biological data is obtained based on the shore image, and hydrological and water quality data is obtained based on the underwater chemical composition, the underwater water flow information and the surface image.

[0011] Preferably, the steps of constructing a water area twin map based on the geographic information, adding the biodiversity database and the hydrological and water quality database to the water area twin map, and generating a visual ecological environment supervision map are specifically as follows: Extracting the water body ecological range and the shore ecological range based on the geographic information, and constructing a water area twin map based on the water body ecological range and the shore ecological range; According to the biodiversity database, simulate and generate multiple biological species and determine the number of organisms and main living areas of the biological species; Displaying the species, number and main living areas of the organisms in the water twin map; According to the hydrological and water quality database, the approximate water flow direction, water quality and hydrological data, and pollution data of each area in each water area are obtained; The approximate water flow direction, the water quality and hydrological data, and the pollution data are displayed in the water area twin map, and the pollution data is marked in red as a warning.

[0012] In summary, this application includes at least one of the following beneficial technical effects: By acquiring geographic information about a water area, multiple water quality monitoring points are established within the area. These monitoring points monitor past water quality and collect pollution data when changes in water quality or pollution are detected. Hydrological models, meteorological data, and data on surrounding enterprises are then obtained. Water flow patterns are determined based on the hydrological models and meteorological data, and the potential pollution emissions from each enterprise are determined based on the surrounding enterprise data. These data are then compared with the pollution data to identify suspected target enterprises. The pollution paths are then deduced to identify the suspected sources of pollution. Underwater robots and drones are then deployed to survey the suspected pollution sources, generating survey data. Based on this data, survey reports are generated and reported to regulatory authorities. Drones and underwater robots are then deployed to conduct regular patrols of the water area, collecting underwater and shoreline biological data as well as hydrological and water quality data to establish biodiversity and hydrological and water quality databases, respectively. A twin map of the water area is then constructed based on the geographic information of the water area, and the biodiversity and hydrological and water quality databases are visualized on the twin map. This improves the efficiency and real-time nature of water ecological and environmental supervision. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 This is a module block diagram of an artificial intelligence-based water ecological environment supervision system provided in an embodiment of the present application.

[0014] Explanation of the accompanying symbols: 1. Pollution monitoring module; 2. Inverse deduction module; 3. Investigation and reporting module; 4. Ecological collection module; 5. Ecological visualization module. DETAILED DESCRIPTION

[0015] The following combination Figure 1 This application is further described in detail, but the embodiments of the present invention are not limited thereto.

[0016] The embodiment of the present application discloses an artificial intelligence-based water ecological environment monitoring system.

[0017] In this embodiment, a water ecological environment monitoring system based on artificial intelligence includes: Pollution monitoring module 1: used to obtain geographical information of the water area, establish multiple water quality monitoring points based on the geographical information, obtain water quality data, and obtain pollution data based on the water quality data; Inverse deduction module 2: used to obtain the hydrological model, meteorological data and surrounding enterprise data of the water area, and perform inverse deduction of the pollution path based on the hydrological model, pollution data, meteorological data and surrounding enterprise data to determine the suspected pollution source; Investigation and reporting module 3: It is used to dispatch underwater robots and drones to investigate suspected pollution sources, obtain investigation data, identify pollution source enterprises based on the investigation data, generate an investigation report, and send the investigation report to the regulatory authorities; Ecological collection module 4: used to call underwater robots and drones to conduct regular patrols in the water area, and collect underwater biological data, shore biological data, and hydrological and water quality data during the patrol process. The biodiversity database is constructed based on the underwater biological data and shore biological data, and the hydrological and water quality database is constructed based on the hydrological and water quality data. Ecological visualization module 5: used to construct a water area twin map based on geographic information, add the biodiversity database and hydrological and water quality database to the water area twin map, and generate a visual ecological environment supervision map.

[0018] It should be noted that the above modules are only basic modules of this embodiment. During the specific implementation process, some modules can be appropriately added, reduced or modified without affecting the overall implementation effect.

[0019] The steps of obtaining geographic information of the water area, establishing multiple water quality monitoring points based on the geographic information, obtaining water quality data, and obtaining pollution data based on the water quality data are as follows: Obtain geographic information, and obtain the main water area, number of tributaries, and length of tributaries based on the geographic information; According to the scope of the main water area, a uniform monitoring network is established to patrol the main water area and identify high-risk areas that need to be monitored; According to the high-risk areas, high-risk monitoring points are determined, and the uniform monitoring network is adjusted according to the high-risk monitoring points to generate a monitoring network; Extract the intersection points of the monitoring network and establish water quality monitoring points within the main water area based on the intersection points; According to the number and length of tributaries, water quality monitoring points are established on each tributary at preset length intervals; Each water quality monitoring point is called to monitor the water quality in the water area, obtain water quality data, and obtain pollution data based on the water quality data.

[0020] In practice, a 15-kilometer-long Dongjiang River tributary within a certain municipal district was used as an example. When constructing a monitoring network, geographic information obtained for this area revealed a main channel width of 78 meters, an average depth of 2.8 meters, and five tributaries, each 1.2 kilometers, 3.5 kilometers, 0.9 kilometers, 2.1 kilometers, and 4.3 kilometers long. Remote sensing mapping generated a 20×20 meter grid monitoring layout, revealing an excessive concentration of heavy metal cadmium (0.015 mg / L, twice the Class III standard) 1.2 kilometers downstream of an industrial sewage outfall. Based on risk modeling, the density of monitoring points in this high-risk area was increased from one every 50 meters to one every 10 meters, with eight additional multi-parameter water quality sensors installed. Monitoring points were also deployed every 300 meters along the tributary, with 15 buoy-type monitoring stations deployed along the longest 4.3-kilometer-long tributary. During actual operation, monitoring point DJ-09 detected a sudden increase in ammonia nitrogen concentration to 2.8 mg / L (4.7 times the standard), accompanied by a sharp drop in pH from 7.2 to 6.1. The system, integrating data from 12 upstream and downstream monitoring points, confirmed the contamination zone to be a ribbon-like spread, covering 800 meters in length and a maximum width of 60 meters. The source was traced back to the drainage outlet of an electroplating plant.

[0021] Obtain the hydrological model, meteorological data, and surrounding enterprise data of the water area, and reversely deduce the pollution path based on the hydrological model, pollution data, meteorological data, and surrounding enterprise data to determine the suspected pollution source. The specific steps are as follows: Obtain the hydrological model, meteorological data, and surrounding enterprise data of the water area, and determine the direction of water flow in the water area based on the hydrological model and meteorological data; Based on the data of surrounding enterprises, the types of waste generated by surrounding enterprises in their production processes are obtained, and the chemical composition of emissions is obtained based on the types of waste; Extract the pollution data to obtain the target monitoring location and pollution chemical data in the pollution data; Determine the approximate direction and range of the pollution source based on the direction of water flow and target monitoring location; Compare the pollution chemical data with the emission chemical composition to obtain a comparison value, and screen multiple surrounding enterprises based on the comparison value to obtain multiple suspected target enterprises; Based on the general direction and range and multiple suspected target enterprises, the pollution path is reversed to determine the suspected pollution source.

[0022] During application, the system retrieved a stored hydrological model for a benzene pollution incident in a section of a Dongjiang River tributary, revealing a normal-water flow velocity of 0.25 m / s, with a predominant southwesterly direction. This data, combined with the meteorological station's 48-hour cumulative rainfall of 38 mm and a wind speed of level 2 southeasterly, showed that monitoring point DJ-14 detected benzene concentrations of 1.05 mg / L (21 times the Class V standard) and toluene of 0.78 mg / L (15.6 times the Class V standard). The enterprise database indicated the presence of Chemical Plant B (registered for a benzene emission permit) and Paint Plant C (undeclared any related pollutants) within 3 kilometers upstream. A reverse flow model calculated the pollution source to be located 1.5-2.2 kilometers northeast of the monitoring point, coinciding with the coordinates of Plant B (1.8 kilometers from the center of the contamination zone) and Plant C (2.1 kilometers from the center of the contamination zone). Mass spectrometry analysis revealed a 92% match between the characteristic peaks of benzene series in the contaminated water and those in wastewater samples from Plant B, while no similar compounds were detected in samples from Plant C. Ultimately, a pollution contribution rate algorithm determined that Plant B was the sole source of pollution, with failures in its wastewater treatment facilities resulting in substandard discharges.

[0023] After determining the suspected pollution source by reverse deducing the pollution path based on the general direction and scope and multiple suspected target enterprises, the following steps are also included: If no suspected target enterprise is found after screening multiple surrounding enterprises based on the comparison value, the possible pollution source type is obtained based on the geographical information and the approximate direction range; According to the possible pollution source type, the formation mode of the pollution source is further obtained, and the scope of the pollution source is determined according to the formation mode; Based on geographic information and the scope of pollution sources, reverse deduction of pollution paths is carried out, searches are conducted within the scope of pollution sources, and suspected pollution sources are identified.

[0024] During an algal bloom event in a bend of a Dongjiang River tributary, the system detected an abnormal rise in chlorophyll a concentration to 28 μg / L (1.8 times the Class III standard), with dissolved oxygen fluctuating between 4.2 and 9.8 mg / L during the day and night. Geographic information analysis revealed the pollution core area was located 3.6 kilometers north of the tributary. Within a 5-kilometer radius, there were no industrial enterprises, but 12 vegetable greenhouses existed. Satellite imagery interpretation revealed a contiguous area of 380 mu (approximately 16 acres) of farmland 2.8 kilometers upstream. Fertilization records indicated the application of 50 kg of compound fertilizer per mu over the past two weeks. The pollution source database matched agricultural non-point source pollution with a total nitrogen loss coefficient of 0.18 and a total phosphorus loss coefficient of 0.12. Topographic analysis determined a catchment area of 4.2 square kilometers. With a soil permeability of 0.15 cm / min, the pollution spread was calculated to be a sector with a diameter of 950 meters. The field sampling robot measured 9.6 mg / L of total nitrogen and 1.2 mg / L of total phosphorus from the farmland drainage ditch, which showed a significant positive correlation with the river pollution data (R²=0.87), confirming that eutrophication was caused by excessive application of chemical fertilizers.

[0025] The steps for dispatching underwater robots and drones to survey suspected pollution sources, obtain survey data, identify pollution source enterprises based on the survey data, and generate a survey report are as follows: Construct underwater inspection routes and aerial inspection routes based on geographic information; Dispatch underwater robots to conduct surveys according to the underwater inspection route and obtain underwater survey data; Dispatch drones to conduct surveys according to aerial inspection routes and obtain aerial survey data; Obtaining a first spatiotemporal scale of the underwater robot and a second spatiotemporal scale of the drone, and aligning the first spatiotemporal scale with the second spatiotemporal scale when the underwater robot and the drone conduct a survey; Align the underwater survey data and the aerial survey data to generate an underwater-aerial linkage data table; Based on the underwater-air linkage data table, pollution source tracing is carried out to obtain the pollution source enterprises and generate an investigation report.

[0026] In practice, during an oil spill at a cargo terminal on a Dongjiang River tributary, the system planned a three-dimensional patrol route consisting of a 2.8-kilometer underwater route and a 4.5-kilometer aerial route. The deployed Dongjiang-3 underwater robot, equipped with a fluorescence sensor, detected a 1.5-cm-thick diesel oil slick at coordinates E114°22'16"N23°05'28". Chromatographic analysis confirmed a cetane number of 48.5. A synchronized "Xunjiang-7" unmanned aerial vehicle (UAV) identified a 1.2-kilometer-long oil spill using hyperspectral imaging. Its thermal infrared sensor indicated that the oil slick's temperature was 2.3°C higher than the surrounding water. A data fusion module superimposed the UAV's 15-cm positioning data with the drone's 3-cm positioning data to generate a pollution distribution map with a spatial error of ≤5 cm. Source tracing, linked to AIS ship track data, pinpointed the cargo vessel Shunjiang 668 as having passed through the area during the pollution period. An 8-meter-long oil stain was visible at the port waterline, consistent with the company's reported 150-liter hydraulic oil leak.

[0027] After constructing underwater inspection routes and aerial inspection routes based on geographic information, the following steps are also included: Based on the aerial inspection route, the route area passed by the aerial inspection route is obtained, and whether there is a sensitive area in the route area is determined; If it is determined that there is a sensitive area within the route area, the inspection route will be changed to bypass the sensitive area by obtaining the supervisory authority of the sensitive area and applying for flight permission from the supervisory authority. According to the route area, determine whether there are any aerial obstacles in the route area; If it is determined that there is an aerial obstacle, the volume information, height information and type information of the aerial obstacle are obtained; Adjust the aerial inspection route based on volume information, altitude information and type information.

[0028] During a flood season inspection of a Dongjiang tributary, the system automatically identified the DJ-12 to DJ-15 monitoring section as a flight control zone. In accordance with Article 28 of the Aviation Control Regulations, a temporary flight application was submitted to the relevant authorities. After obtaining permission, the drone's flight altitude was adjusted from 100 meters to 220 meters, with a speed limit of 8 meters per second. The obstacle avoidance module detected a high-voltage power line across the river (tower height 185 meters, lowest sag point 110 meters above sea level), shifting the original flight path 350 meters northwest and adding two new waypoints. When crossing a cross-river bridge, the millimeter-wave radar measured the 412-meter spacing between the piers in real time, dynamically adjusting the flight attitude to maintain a safe distance of 50 meters from the bridge, and simultaneously activating propeller speed compensation mode to mitigate crosswind interference.

[0029] The steps for respectively calling underwater robots and drones to conduct regular patrols in the water area and collecting underwater biological data, shore biological data, and hydrological and water quality data during the patrols are as follows: Based on the uniform monitoring network, the main water patrol routes of underwater robots and drones are constructed. The tributary water patrol routes are generated according to the number and length of tributaries. The water patrol routes are generated according to the main water patrol routes and tributary water patrol routes. Bind an underwater robot and a drone to form a patrol group. Release a patrol group at preset time intervals to conduct regular patrols along the water patrol route. Based on geographic information, the edge outline of the water area is obtained, and the ecological range of the shore is obtained by expanding outward based on the edge outline; Generate shore patrol routes based on the ecological range of the shore, and release a drone at preset time intervals to conduct regular patrols along the shore patrol routes; During patrols, underwater robots collect underwater images, underwater chemical composition, and underwater water flow information, while drones collect surface and shore images. Underwater biological data is obtained based on underwater images and surface images, shore biological data is obtained based on shore images, and hydrological and water quality data is obtained based on underwater chemical composition, underwater water flow information and surface images.

[0030] In practice, during the monitoring of the Dongjiang River tributary ecological restoration project, the system planned patrol routes consisting of 18 longitudinal and 9 transverse lines, forming an 800×800 meter monitoring grid. Three teams of "Hujiang-5" underwater robots were deployed, each equipped with two "Kanjiang-II" drones, to conduct four full-basin scans daily. The ecological protection zone demarcation algorithm extended the riverbank 300 meters inland, generating a total of 42 kilometers of key patrol routes. During the spring 2024 monitoring, the robot DJ-EC03 recorded nine sonar signals from the Chinese sturgeon, confirming that its migration path is located in water depths of 4 meters or more on the west side of the tributary. The drone's multispectral camera detected an increase in submerged plant cover from 12% to 27%. The water quality sensor network showed a decrease in ammonia nitrogen concentration from 0.58 mg / L to 0.21 mg / L, while dissolved oxygen stabilized between 7.2 and 8.5 mg / L, indicating an improvement in the water quality from Class IV to Class III standards.

[0031] Based on geographic information, a water area twin map is constructed, and the biodiversity database and the hydrological and water quality database are added to the water area twin map to generate a visual ecological environment supervision map. The specific steps are as follows: Based on geographic information, the ecological range of the water body and the ecological range of the shore are extracted, and a twin map of the water area is constructed based on the ecological range of the water body and the ecological range of the shore; Based on the biodiversity database, simulate and generate a variety of biological species and determine the number of species and their main habitats; Display the species, number of organisms and main living areas in the water twin map; According to the hydrological and water quality database, the approximate water flow direction, water quality and hydrological data, and pollution data of each area in each water area are obtained; The approximate water flow direction, water quality and hydrological data, and pollution data are displayed on the water area twin map, and the pollution data is marked in red as a warning.

[0032] In practice, the digital twin system for the Dongjiang River tributaries was constructed, integrating ecological data covering 85 square kilometers of water, including 3D models of 32 fish species and 18 benthic organisms. The hydrological visualization layer shows an average flow velocity of 0.32 m / s in the main channel, with a vortex zone 120 meters in diameter forming at the confluence of the tributaries. The pollution warning module dynamically annotated the lead concentration of 0.018 mg / L (1.2 times the Class III standard) detected at monitoring point DJ-07, and linked it to heavy metal emission records from five upstream enterprises over the past three months. Supervisors can access a 72-hour water quality curve for any section through a touchscreen interface, such as the COD index at section DJ-15 dropping from 32 mg / L to 18 mg / L, visually demonstrating the effectiveness of sewage treatment plant upgrades. The emergency simulation module predicts that if a 20-ton diesel spill occurs, the contamination zone will affect an area 12 kilometers downstream within six hours, guiding the development of a Level 3 emergency response plan.

[0033] The above are all preferred embodiments of the present application, and are not intended to limit the scope of protection of the present application. Therefore, any equivalent changes made based on the structure, shape, and principle of the present application should be included in the scope of protection of the present application.

Claims

1. An artificial intelligence-based water ecological environment monitoring system, characterized by: The system includes: Pollution monitoring module: used to obtain geographical information of the water area, establish multiple water quality monitoring points based on the geographical information, obtain water quality data, and obtain pollution data based on the water quality data; Inverse deduction module: used to obtain the hydrological model, meteorological data and surrounding enterprise data of the water area, and perform inverse deduction of the pollution path based on the hydrological model, the pollution data, the meteorological data and the surrounding enterprise data to determine the suspected pollution source; Investigation and reporting module: used to dispatch underwater robots and drones to investigate the suspected pollution sources, obtain investigation data, identify the pollution source enterprises based on the investigation data, generate an investigation report, and send the investigation report to the regulatory authorities; Ecological collection module: used to call underwater robots and drones to conduct regular patrols in the water area, and collect underwater biological data, shore biological data, and hydrological and water quality data during the patrol process, build a biodiversity database based on the underwater biological data and the shore biological data, and build a hydrological and water quality database based on the hydrological and water quality data; Ecological visualization module: used to construct a water area twin map based on the geographic information, add the biodiversity database and the hydrological and water quality database to the water area twin map, and generate a visual ecological environment supervision map.

2. The artificial intelligence-based water ecological environment monitoring system according to claim 1 is characterized in that: The steps of obtaining geographic information of a water area, establishing multiple water quality monitoring points based on the geographic information, obtaining water quality data, and obtaining pollution data based on the water quality data are specifically as follows: Obtaining geographic information, and obtaining the main water area, the number of tributaries, and the length of the tributaries according to the geographic information; Based on the scope of the main water area, a uniform monitoring network is constructed to patrol the main water area and identify high-risk areas that require key monitoring; Determining high-risk monitoring points based on the high-risk areas, and adjusting the uniform monitoring network based on the high-risk monitoring points to generate a monitoring network; Extracting the intersection positions of the network intersections of the monitoring network, and establishing water quality monitoring points within the main water area according to the intersection positions; According to the number and length of the tributaries, a water quality monitoring point is established on each tributary at a preset length interval; Each of the water quality monitoring points is called to monitor the water quality in the water area to obtain water quality data, and pollution data is obtained based on the water quality data.

3. The artificial intelligence-based water ecological environment monitoring system according to claim 2 is characterized in that: The steps of obtaining a hydrological model, meteorological data, and data of surrounding enterprises in a water area, and reversely deducing the pollution path based on the hydrological model, the pollution data, the meteorological data, and the data of surrounding enterprises to determine the suspected pollution source are specifically as follows: Obtaining a hydrological model, meteorological data, and surrounding enterprise data of the water area, and obtaining the direction of water flow in the water area based on the hydrological model and the meteorological data; Based on the data of the surrounding enterprises, the types of waste generated by the surrounding enterprises during their production processes are obtained, and the chemical composition of the emissions is obtained based on the types of waste; Extracting the pollution data to obtain target monitoring locations and pollution chemical data in the pollution data; Determine the approximate direction and range of the pollution source based on the water flow direction and the target monitoring location; Comparing the pollution chemical data with the emission chemical composition to obtain a comparison value, and screening the plurality of surrounding enterprises according to the comparison value to obtain a plurality of suspected target enterprises; Based on the general direction range and multiple suspected target enterprises, the pollution path is reversed to determine the suspected pollution source.

4. The water ecological environment monitoring system based on artificial intelligence according to claim 3 is characterized in that: After performing the step of reverse deducing the pollution path based on the general direction range and the multiple suspected target enterprises to determine the suspected pollution source, the method further includes: If no suspected target enterprise is obtained after screening the plurality of surrounding enterprises according to the comparison value, the possible pollution source type is obtained according to the geographic information and the approximate direction range; According to the possible pollution source type, further obtaining the formation mode of the pollution source, and determining the scope of the pollution source according to the formation mode; Based on the geographic information and the scope of the pollution source, the pollution path is reversed, and a search is conducted within the scope of the pollution source to determine the suspected pollution source.

5. The artificial intelligence-based water ecological environment monitoring system according to claim 4 is characterized in that: The steps of dispatching underwater robots and drones to survey the suspected pollution sources, obtaining survey data, determining the pollution source enterprises based on the survey data, and generating a survey report are specifically as follows: constructing underwater inspection routes and aerial inspection routes based on the geographic information; dispatching an underwater robot to conduct an inspection according to the underwater inspection route to obtain underwater inspection data; Dispatching drones to conduct surveys according to the aerial inspection routes to obtain aerial inspection data; Obtaining a first spatiotemporal scale of the underwater robot and a second spatiotemporal scale of the unmanned aerial vehicle, and aligning the first spatiotemporal scale with the second spatiotemporal scale when the underwater robot and the unmanned aerial vehicle conduct a survey; Aligning the underwater survey data and the aerial survey data to generate an underwater-aerial linkage data table; Based on the underwater-air linkage data table, pollution source tracing is carried out to obtain the pollution source enterprises and generate a survey report.

6. The artificial intelligence-based water ecological environment monitoring system according to claim 5 is characterized in that: After constructing the underwater inspection route and the aerial inspection route based on the geographic information, the following steps are further included: Based on the aerial inspection route, obtaining a route area passed by the aerial inspection route, and determining whether there is a sensitive area in the route area; If it is determined that there is a sensitive area in the route area, obtain the regulatory department of the sensitive area, apply for flight permission from the regulatory department, and fly after obtaining the flight permission; otherwise, change the inspection route to bypass the sensitive area; determining, based on the route area, whether there are any aerial obstacles in the route area; If it is determined that there is an aerial obstacle, the volume information, height information, and type information of the aerial obstacle are obtained; The aerial inspection route is adjusted according to the volume information, the altitude information, and the type information.

7. The artificial intelligence-based water ecological environment monitoring system according to claim 6 is characterized in that: The steps for respectively calling underwater robots and drones to conduct regular patrols in the water area and collecting underwater biological data, shore biological data, and hydrological and water quality data during the patrols are as follows: Based on the uniform monitoring network, a main water patrol route for underwater robots and drones is constructed, a tributary water patrol route is generated according to the number of tributaries and the length of the tributaries, and a water patrol route is generated according to the main water patrol route and the tributary water patrol route; Binding an underwater robot and a drone to generate a patrol group, and releasing the patrol group according to a preset time interval to perform regular patrols along the water patrol route; Obtaining the edge outline of the water area based on the geographic information, and expanding outwards based on the edge outline to obtain the shore ecological range; Generate a shore patrol route based on the shore ecological range, and release a drone at preset time intervals to perform regular patrols along the shore patrol route; During patrols, underwater robots collect underwater images, underwater chemical composition, and underwater water flow information, while drones collect surface and shore images. Underwater biological data is obtained based on the underwater image and the surface image, shore biological data is obtained based on the shore image, and hydrological and water quality data is obtained based on the underwater chemical composition, the underwater water flow information and the surface image.

8. The artificial intelligence-based water ecological environment monitoring system according to claim 7 is characterized in that: The steps of constructing a water area twin map based on the geographic information, adding the biodiversity database and the hydrological and water quality database to the water area twin map, and generating a visual ecological environment supervision map are specifically as follows: Extracting the water body ecological range and the shore ecological range based on the geographic information, and constructing a water area twin map based on the water body ecological range and the shore ecological range; According to the biodiversity database, multiple species are simulated and generated, and the number of species and main habitats of the species are determined; Displaying the species, number and main living areas of the organisms in the water twin map; According to the hydrological and water quality database, the approximate water flow direction, water quality and hydrological data, and pollution data of each area in each water area are obtained; The approximate water flow direction, the water quality and hydrological data, and the pollution data are displayed in the water area twin map, and the pollution data is marked in red as a warning.

Images