Groundwater monitoring method and system based on Beidou dual-channel and Bluetooth communication

Abstract

The present application discloses a groundwater monitoring method based on Beidou dual-channel and Bluetooth communication, which relates to the field of groundwater monitoring. The method includes: obtaining surface remote sensing information and meteorological information; calculating groundwater level information based on surface remote sensing information and meteorological information; obtaining surface water area, and generating a monitoring point distribution strategy based on groundwater level information and surface water area; deploying groundwater monitoring stations based on the monitoring point distribution strategy; using groundwater monitoring stations to monitor groundwater; obtaining groundwater monitoring information; constructing a groundwater evaluation model, inputting groundwater monitoring information into the groundwater evaluation model, and obtaining a groundwater evaluation grade; generating a groundwater detection report. The present application can effectively realize high-precision, high-efficiency, and all-round real-time monitoring of groundwater in the target area.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of groundwater monitoring, and in particular, to a groundwater monitoring method and system based on Beidou dual-channel and Bluetooth communication. Background Art

[0002] Groundwater is an important component of the ecological environment and plays a key role in maintaining the stability of the ecosystem and protecting biodiversity. In addition, due to the stable amount and good water quality of groundwater, it is also an important source of daily drinking, agricultural irrigation, industrial and agricultural production. Therefore, groundwater is vital to human society and the ecological environment. However, due to excessive exploitation, pollutant emissions and other reasons, the groundwater level has dropped and the water quality has deteriorated, which in turn threatens human health and safety and the stability of the natural ecosystem. Based on this, in order to timely discover and curb the above situation, it is necessary to conduct high-precision, high-efficiency, and comprehensive real-time monitoring of groundwater.

[0003] Traditional monitoring technology usually involves regular sampling and monitoring of groundwater using portable detection tools to analyze its chemical composition. Although this method can immediately obtain the groundwater quality status of a certain location, it is difficult to conduct comprehensive real-time monitoring of a certain area when there are many monitoring points that need to be monitored in a certain area, and it is also difficult to continuously obtain the water quality status of the area. In order to overcome the shortcomings of manual monitoring, some areas have established a network of monitoring wells in important areas to regularly collect water level and water quality data, but manual analysis of water level and water quality data is still required. Since manual analysis of water level and water quality data is not possible in real time, it may not be possible to timely detect the decline in groundwater level and deterioration of water quality, resulting in serious consequences such as water shortage and harm to human health. At the same time, manual analysis of water level and water quality data will also cause a waste of human resources. Summary of the invention

[0004] The embodiments of the present application provide a groundwater monitoring method and system based on Beidou dual-channel and Bluetooth communication, which is used to solve the problem that the existing technology cannot achieve high-precision, high-efficiency, and comprehensive real-time monitoring of groundwater.

[0005] To achieve the above objectives, the embodiments of the present application adopt the following technical solutions:

[0006] In a first aspect, a groundwater monitoring method based on Beidou dual-channel and Bluetooth communication is provided, the method comprising:

[0007] Obtain surface remote sensing information and meteorological information of the target area through satellite technology;

[0008] Acquire the surface characteristic parameters of the target area based on the surface remote sensing information, assign weights to the surface characteristic parameters according to the meteorological information, and calculate the groundwater level information of the target area using an inversion algorithm;

[0009] Acquire the surface water area of ​​the target area according to the surface remote sensing information, and generate a monitoring point distribution strategy for the target area in combination with the groundwater level information and the surface water area;

[0010] Deploy a plurality of groundwater monitoring stations in the target area according to the monitoring point distribution strategy, wherein the groundwater monitoring stations are provided with an information transmission module, and the information transmission module includes a Beidou communication unit and a Bluetooth communication unit;

[0011] Using the groundwater monitoring station to monitor groundwater in the target area;

[0012] Acquire monitoring information of all the groundwater monitoring stations through the Beidou communication unit and the Bluetooth communication unit, and integrate all the monitoring information to obtain groundwater monitoring information of the target area;

[0013] Constructing a groundwater evaluation model based on a parameter optimization algorithm and a neural network model, inputting the groundwater monitoring information into the trained groundwater evaluation model, and obtaining the groundwater evaluation grade of the target area;

[0014] A groundwater detection report for the target area is generated by combining the groundwater level information, the groundwater monitoring information and the groundwater evaluation grade.

[0015] Optionally, the surface remote sensing information includes optical remote sensing information, microwave remote sensing information and infrared remote sensing information, and the steps of acquiring the surface characteristic parameters of the target area based on the surface remote sensing information, assigning weights to the surface characteristic parameters according to the meteorological information, and calculating the groundwater level information of the target area using an inversion algorithm include the following steps:

[0016] Analyze the band reflectance in the optical remote sensing information to invert the surface vegetation index of the target area;

[0017] Extracting the echo intensity of the microwave signal in the microwave remote sensing information, and calculating the surface soil moisture of the target area according to the echo intensity;

[0018] Acquiring the surface temperature characteristics of the target area through the infrared remote sensing information;

[0019] Integrating the surface vegetation index, the surface soil moisture and the surface temperature characteristics to obtain the surface characteristic parameters of the target area;

[0020] Acquiring regional spatial temperature information and regional precipitation information of the target area according to the meteorological information, and assigning weights to the surface characteristic parameters according to the regional spatial temperature information and the regional precipitation information;

[0021] The groundwater level information of the target area is calculated based on the surface vegetation index, the surface soil moisture and the surface temperature after weight distribution and using an inversion algorithm.

[0022] Optionally, the surface remote sensing information also includes a surface remote sensing image, and the step of acquiring the surface water distribution characteristics of the target area according to the surface remote sensing information and generating a monitoring point distribution strategy for the target area in combination with the groundwater level information and the surface water distribution characteristics includes the following steps:

[0023] Extracting optical reflection characteristics of the target area through the optical remote sensing information, and marking the surface water area of ​​the target area according to the optical reflection characteristics;

[0024] Dividing the target area into grids according to the surface remote sensing image to obtain a plurality of grid areas;

[0025] For each of the grid areas of the target area, calculating the area of ​​the surface water region within the grid area to obtain the surface water area;

[0026] The number of monitoring points in each grid area is set in combination with the surface water area and the groundwater level information corresponding to the grid area, and the location of the monitoring points in the grid area is generated according to the number of monitoring points;

[0027] The monitoring point locations of all the grid areas are integrated to obtain a monitoring point distribution strategy for the target area.

[0028] Optionally, acquiring the monitoring information of all the groundwater monitoring stations through the Beidou communication unit and the Bluetooth communication unit, and integrating all the monitoring information to obtain the groundwater monitoring information of the target area comprises the following steps:

[0029] Acquire the first groundwater information collected by all the groundwater monitoring stations through the Beidou communication unit;

[0030] Matching the first groundwater information with the corresponding monitoring point locations, and marking the monitoring point locations where information is missing as abnormal monitoring points;

[0031] For each of the grid areas, calculating the information difference between all the first groundwater information within the grid area;

[0032] Screening out the first groundwater information whose information difference is greater than a preset difference threshold, and marking the monitoring point position corresponding to the first groundwater information as the abnormal monitoring point;

[0033] According to the monitoring point position corresponding to the abnormal monitoring point and using the shortest path algorithm, an information collection route of a monitoring drone is generated, wherein the monitoring drone is equipped with a Bluetooth communication module;

[0034] Control the monitoring drone to obtain the second groundwater information of all the abnormal monitoring points through the Bluetooth communication unit according to the information collection route;

[0035] The first groundwater information after information screening is integrated with the second groundwater to obtain the groundwater monitoring information of the target area.

[0036] Optionally, constructing a groundwater evaluation model based on a parameter optimization algorithm and a neural network model, inputting the groundwater monitoring information into the trained groundwater evaluation model, and obtaining a groundwater evaluation grade comprises the following steps:

[0037] Using wavelet transform to eliminate abnormal information in the groundwater monitoring information, and calculating the information gain of each water quality attribute information in the groundwater monitoring information after the wavelet transform, and screening out the core water quality information in the water quality attribute information according to the information gain;

[0038] A primary groundwater assessment model was constructed based on the convolutional neural network model and the long short-term memory network model;

[0039] Using a pre-constructed water quality training set to perform model training on the primary groundwater assessment model;

[0040] During the model training process, the optimal model parameters of the primary groundwater assessment model are calculated using a parameter optimization algorithm;

[0041] Optimizing the parameters of the primary groundwater assessment model according to the optimal model parameters to obtain a groundwater assessment model;

[0042] The core water quality information is input into the trained groundwater evaluation model to obtain a groundwater evaluation grade.

[0043] Optionally, the step of eliminating abnormal information in the groundwater monitoring information by wavelet transform, calculating the importance of water quality attributes of each water quality attribute information in the groundwater monitoring information after the wavelet transform, and selecting core water quality information in the water quality attribute information according to the importance of the water quality attributes includes the following steps:

[0044] Using wavelet basis functions to perform multi-scale decomposition on the groundwater monitoring information to obtain multiple groundwater monitoring information components and high-frequency coefficients corresponding to the groundwater monitoring information components;

[0045] Eliminate the groundwater monitoring information components whose high-frequency coefficients are less than a preset coefficient threshold in the groundwater monitoring information components;

[0046] Reconstructing the groundwater monitoring information components that have been eliminated to obtain normal groundwater monitoring information;

[0047] Calculating the information entropy of each item of water quality attribute information in the normal groundwater monitoring information, and calculating the information gain of the corresponding water quality attribute information according to the information entropy;

[0048] All the water quality attribute information whose information gain is greater than a preset gain threshold in the water quality attribute information are screened out as core water quality information.

[0049] Optionally, the step of calculating the optimal model parameters of the primary groundwater assessment model using a parameter optimization algorithm comprises the following steps:

[0050] constructing an initial population of a parameter optimization algorithm using a plurality of parameters to be optimized of the primary groundwater assessment model;

[0051] Initializing the initial population, and inputting the initialized initial population into the primary groundwater evaluation model to obtain an initial output value of the primary groundwater evaluation model;

[0052] Calculating a root mean square error between the initial output value and a pre-acquired reference output value, and using the root mean square error as the fitness of the initial population;

[0053] Retaining the population individuals whose fitness in the initial population is greater than a preset first fitness threshold to obtain a core population;

[0054] Taking the population individuals whose fitness in the core population is greater than a preset second fitness threshold as core individuals, where the second fitness threshold is greater than the first fitness threshold;

[0055] Based on the core individuals and using a dynamic weight coefficient strategy, the core population is updated to a target core population;

[0056] Using the Levy flight strategy to update and adjust the core individual with the best fitness in the target core population, and calculating the updated target fitness of all population individuals in the target core population;

[0057] The population updating iteration step is repeated for the target core population until a preset maximum number of iterations is reached, thereby obtaining the optimal model parameters of the primary groundwater assessment model.

[0058] Optionally, inputting the core water quality information into the trained groundwater evaluation model to obtain the groundwater evaluation grade of the target area comprises the following steps:

[0059] Inputting the core water quality information into the spatial feature extraction module of the groundwater assessment model to extract spatial features, thereby obtaining local groundwater features of the target area, wherein the spatial feature extraction module is constructed based on a convolutional neural network model;

[0060] Inputting the core water quality information into the time series feature extraction module of the groundwater assessment model for time series feature analysis to obtain the global groundwater characteristics of the target area within a preset time period, wherein the time series feature extraction module is constructed based on a long short-term memory neural network model;

[0061] Inputting the local groundwater characteristics and the global groundwater characteristics into the multi-scale feature fusion module of the groundwater assessment model for feature fusion to obtain the groundwater fusion characteristics of the target area;

[0062] After the groundwater fusion features are input into the fully connected layer of the groundwater evaluation model, the groundwater level of the target area is evaluated using a preset classification function to obtain the groundwater evaluation level of the target area.

[0063] Optionally, the generating of the groundwater detection report of the target area by combining the groundwater level information, the groundwater monitoring information and the groundwater evaluation grade comprises the following steps:

[0064] Extracting water level information of monitoring points from the groundwater monitoring information;

[0065] Fusing the monitoring point water level information with the groundwater level information to obtain global groundwater level information of the target area;

[0066] A groundwater detection report for the target area is generated by combining the global groundwater level information, groundwater monitoring information and the groundwater evaluation grade.

[0067] In the second aspect, the present application discloses a groundwater monitoring system based on Beidou dual-channel and Bluetooth communication, characterized in that it includes:

[0068] a memory configured to store instructions; and

[0069] The processor is configured to call the instruction from the memory and implement a method for groundwater monitoring based on Beidou dual-channel and Bluetooth communication according to any one of the first aspects when executing the instruction.

[0070] Through the above technical scheme, the groundwater level information is inverted through the surface remote sensing information, and the monitoring point distribution strategy of the target area is designed according to the groundwater level information and the surface water area, and multiple groundwater monitoring stations are deployed in the target area according to the monitoring point distribution strategy, so as to realize the comprehensive monitoring of the groundwater in the target area. At the same time, different numbers of monitoring points are set for areas with different groundwater information and surface area in the target area, which can save the resource investment for groundwater monitoring and reduce the calculation amount of information analysis. Afterwards, the groundwater monitoring information of the groundwater monitoring station is obtained by using the two different communication methods of Beidou communication unit and Bluetooth communication unit, which can ensure the integrity of the groundwater monitoring information. Finally, the groundwater evaluation model is constructed based on the neural network model and the parameters are optimized by the parameter optimization algorithm, which can make the output result of the groundwater evaluation model more accurate, the obtained groundwater evaluation grade more reliable, and the groundwater evaluation model can analyze the groundwater monitoring information of the target area in real time, and ensure the specific timeliness of the obtained groundwater evaluation grade. In summary, the present application provides a method for real-time monitoring of groundwater with high precision, high efficiency and all-round.

[0071] Other features and advantages of the embodiments of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 A schematic diagram of a process for groundwater monitoring based on Beidou dual-channel and Bluetooth communication provided in an embodiment of the present application;

[0073] Figure 2 A schematic diagram of the information collection route and abnormal monitoring point distribution of the monitoring drone provided in the embodiment of the present application;

[0074] Figure 3 A schematic diagram of a process for optimizing model parameters based on a parameter optimization algorithm provided in an embodiment of the present application. DETAILED DESCRIPTION

[0075] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the specific implementation methods described herein are only used to illustrate and explain the embodiments of the present application, and are not used to limit the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0076] It should be noted that if the embodiments of the present application involve directional indications (such as up, down, left, right, front, back...), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0077] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present application, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0078] Figure 1 The flowchart of a method for groundwater monitoring based on Beidou dual-channel and Bluetooth communication according to an embodiment of the present application is schematically shown. Figure 1 As shown, the embodiment of the present application provides a method for groundwater monitoring based on Beidou dual-channel and Bluetooth communication, which may include the following steps:

[0079] S101. Obtain surface remote sensing information and meteorological information of the target area through satellite technology.

[0080] In this embodiment, satellite technology refers to the use of various types of satellites to obtain surface remote sensing information and meteorological information of the target area. For example, a meteorological satellite can obtain regional spatial temperature information and regional precipitation information of the target area, that is, the meteorological information of the target area. Remote sensing satellites capable of obtaining surface remote sensing information include optical remote sensing satellites equipped with high-resolution optical sensors, which can obtain optical remote sensing information of the target area by capturing and analyzing electromagnetic wave signals such as visible light and infrared light reflected, scattered and emitted by the surface; microwave remote sensing satellites equipped with microwave sensors can record and analyze microwave radiation signals on the surface or in the atmosphere to obtain microwave remote sensing information of the target area; and thermal infrared remote sensing satellites capable of using thermal infrared sensors to obtain infrared remote sensing information of the target area.

[0081] S102, acquiring surface characteristic parameters of the target area based on surface remote sensing information, assigning weights to the surface characteristic parameters according to meteorological information, and calculating groundwater level information of the target area using an inversion algorithm.

[0082] In this embodiment, the surface remote sensing information includes optical remote sensing information, microwave remote sensing information and infrared remote sensing information, and the surface characteristic parameters refer to the surface vegetation index, surface soil moisture and surface temperature. The vegetation growth condition is quantified by analyzing the reflectivity difference between the near infrared band (NIR) and the red light band (RED) in the optical remote sensing information to obtain the vegetation index (NDVI), and the calculation formula is NDVI = (NIR-RED) / (NIR+RED). The surface soil moisture is calculated by analyzing the echo intensity of the microwave signal through microwave remote sensing information, extracting the dielectric constant of the soil, and then calculating the surface soil moisture of the target area. The infrared remote sensing information is radiated and converted into radiance. The atmospheric correction model (such as MODTRAN, SBDART) is used to correct the atmospheric influence to obtain the surface radiance, and then the surface radiance is converted into surface temperature characteristics according to Planck's blackbody radiation law. After obtaining the surface vegetation index, surface soil moisture and surface temperature characteristics, weights are assigned to them according to the meteorological information of the target area. This is because the reliability of the surface vegetation index, surface soil moisture and surface temperature characteristics is different under different climatic conditions. For example, in humid areas with high atmospheric precipitation, precipitation will change the capillary water distribution of the soil profile. Only after the water infiltrated by atmospheric precipitation evaporates can it be inverted. Due to continuous precipitation, atmospheric precipitation cannot evaporate. Therefore, it is not accurate to invert the groundwater level through surface soil moisture. At this time, the weight of surface soil moisture needs to be reduced. Although the surface temperature characteristics can also be applied in humid areas with high atmospheric precipitation, the inversion accuracy will be reduced under low temperature conditions such as winter. This is because groundwater affects the temperature of the surface soil through heat conduction effect and evaporation heat absorption effect, and the heat source effect of groundwater in winter will interfere with the evaporative cooling effect and thus affect the inversion accuracy of the surface temperature characteristics. After completing the weight allocation, the groundwater level information of the target area is inverted by combining the surface vegetation index, surface soil moisture and surface temperature characteristics. Specifically, the surface vegetation index, surface soil moisture and surface temperature characteristics are input into the groundwater inversion model constructed in advance according to the geological conditions and climatic conditions of the target area to obtain the groundwater level information of the target area.

[0083] S103, obtaining the surface water area of ​​the target area according to the surface remote sensing information, and generating a monitoring point distribution strategy for the target area in combination with the groundwater level information and the surface water area.

[0084] In this embodiment, the surface water area refers to the area where rivers and streams and other water bodies are distributed on the surface. When visible light, near infrared light and other light are irradiated on the surface of the water body, reflection, scattering and absorption phenomena will occur. For example, the reflectivity of the blue-green band in the visible light band is higher than that of the red band, while the reflection energy of the near infrared and short-wave infrared parts is less. This is significantly different from the spectral characteristics of vegetation, which helps to distinguish between water bodies and land. After the surface water body is identified, it is marked in the surface remote sensing image. Then, the target area is gridded according to the surface remote sensing image of the target area to obtain multiple grid areas. The area measurement tool in the geographic information system software or remote sensing processing software can be used to calculate the area of ​​the water body marked in each grid area in the surface remote sensing image to obtain the surface water area of ​​each grid area. The number of monitoring points in each grid area is set according to the groundwater level information corresponding to the surface water area and the grid area, and the corresponding number of groundwater monitoring points are set in each grid area according to the number of monitoring points, which can be evenly distributed in the grid area. The monitoring point distribution strategy of the target area is obtained by integrating the positions of all monitoring points in all grid areas. Setting up a monitoring point distribution strategy based on groundwater levels and surface water distribution can maximize the guarantee of comprehensive monitoring of groundwater in the target area, ensure that certain key areas are not missed, and ensure that the final monitoring results are more accurate. At the same time, for some unimportant areas, reducing the distribution of monitoring points can save certain resources.

[0085] S104. Deploy multiple groundwater monitoring stations in the target area according to the monitoring point distribution strategy. The groundwater monitoring stations are provided with an information transmission module, which includes a Beidou communication unit and a Bluetooth communication unit.

[0086] In this embodiment, groundwater monitoring stations are deployed at the monitoring point locations corresponding to the target area according to the monitoring point distribution strategy. The groundwater monitoring stations are provided with an information transmission module, an information collection module, and an information positioning module. The information transmission module includes a Beidou communication unit and a Bluetooth communication unit. In most cases, the collected groundwater monitoring information is transmitted to the groundwater monitoring platform through satellite communication via the Beidou communication unit. If it is located in a place with complex terrain such as a mountainous area or a forest, or if there is a satellite communication device failure or insufficient power in the satellite communication device, the satellite signal may be unstable or unable to establish a connection, thereby making it impossible to successfully transmit the collected groundwater monitoring information. At this time, it is necessary to use a spare Bluetooth communication unit for information transmission. Therefore, it is necessary to set up two communication units to ensure that the groundwater monitoring information collected by the groundwater monitoring station can be successfully transmitted.

[0087] The information collection module includes various types of groundwater monitoring equipment, such as water level monitors, water temperature monitors, water quality monitors, etc., which can monitor various parameters of groundwater, including but not limited to pH value, dissolved oxygen, conductivity, water level, water temperature, turbidity, etc. The information collection module can obtain information of various dimensions of groundwater, which is convenient for subsequent comprehensive evaluation of groundwater quality. The information positioning module locates the groundwater monitoring station through Beidou satellites. When a groundwater monitoring station sends groundwater monitoring information to the groundwater monitoring platform, it also sends the monitoring station location information (monitoring point location) of the groundwater monitoring station, so that the groundwater monitoring information collected by the groundwater monitoring station can be matched with the groundwater monitoring point according to the monitoring station location information, so as to timely discover the source of abnormal groundwater monitoring information, and collect information or troubleshoot the groundwater monitoring station that monitors abnormal data. For example, if the turbidity is missing from the groundwater monitoring information transmitted from a groundwater monitoring station to the groundwater monitoring platform, the groundwater monitoring platform can find the corresponding groundwater monitoring station based on the location information of the groundwater monitoring station sent by it, and can subsequently use the monitoring drone to re-collect the groundwater monitoring information of the groundwater monitoring station. If the turbidity information is still missing, it is determined that the turbidity meter of the groundwater monitoring station is faulty, and relevant personnel will be immediately dispatched to the groundwater monitoring station for equipment maintenance or replacement based on the location information of the groundwater monitoring station, which greatly improves the efficiency of groundwater monitoring equipment maintenance. At the same time, groundwater monitoring can be re-performed using the replaced or repaired equipment to ensure that the groundwater monitoring information is more reliable and comprehensive.

[0088] S105. Use groundwater monitoring stations to monitor groundwater in target areas.

[0089] In this embodiment, the groundwater monitoring equipment in the groundwater monitoring station is used to monitor the groundwater in the target area in real time, including the pH value, dissolved oxygen, conductivity, water level, water temperature, turbidity, etc. of the groundwater in the target area. In addition, if there is an industrial area within a preset range around the monitoring point in the target area, heavy metals (such as mercury, cadmium, chromium, etc.) and organic pollutants (such as petroleum, volatile phenols, etc.) will also be monitored.

[0090] S106. Acquire monitoring information of all groundwater monitoring stations through the Beidou communication unit and the Bluetooth communication unit, and integrate all monitoring information to obtain groundwater monitoring information of the target area.

[0091] In this embodiment, the first groundwater information collected by the groundwater monitoring station is first transmitted to the groundwater monitoring platform through the Beidou communication unit, and the received first groundwater information is matched with the corresponding monitoring point position. When there is a vacancy in the information, the monitoring point position corresponding to the groundwater monitoring information with the vacancy is marked as an abnormal monitoring point. Then the information difference between the first groundwater information in each grid area is calculated, and the first groundwater information with an information difference greater than the preset difference threshold is screened out, and the corresponding monitoring point position is marked as an abnormal monitoring point. After marking all abnormal monitoring points, the information collection route of the monitoring drone is generated according to the position of the abnormal monitoring point using the shortest path algorithm, and the monitoring drone is controlled to go to the abnormal monitoring point according to the information collection route, and the second groundwater information transmitted by the groundwater monitoring station through the Bluetooth communication unit is obtained by using the Bluetooth communication module loaded on it. After the abnormal information in the first groundwater information is eliminated, it is integrated with the second groundwater information to obtain the complete monitoring information of the target area, that is, the groundwater monitoring information. By obtaining groundwater monitoring information twice, abnormal data caused by abnormalities in groundwater monitoring equipment or abnormal transmission of communication equipment can be eliminated to the greatest extent, ensuring that the groundwater monitoring information is more accurate and the subsequent groundwater evaluation results based on the groundwater monitoring information are more reliable.

[0092] S107. Construct a groundwater evaluation model based on the parameter optimization algorithm and the neural network model, input the groundwater monitoring information into the trained groundwater evaluation model, and obtain the groundwater evaluation level of the target area.

[0093] In this embodiment, wavelet transform is used to remove abnormal information in the groundwater monitoring information again, and then the information gain of each water quality attribute information in the groundwater monitoring information is calculated. The core water quality information in the water quality attribute information is screened out according to the information gain. Through this method, the abnormal change data of the groundwater monitoring information in the time series can be eliminated. By calculating the information gain of each water quality attribute information in the groundwater monitoring information, it is possible to screen out a number of water quality attribute information that has a greater impact on the groundwater quality, thereby simplifying the groundwater monitoring information, improving the efficiency of subsequent groundwater evaluation, and reducing the amount of calculation. Real-time groundwater monitoring and rapid analysis of groundwater monitoring information can be carried out to achieve real-time grasp of groundwater water quality and water level changes.

[0094] A primary groundwater evaluation model is constructed based on the convolutional neural network model and the long short-term memory network model. The primary groundwater evaluation model is trained using the pre-constructed water quality training set. At the same time, the optimal model parameters of the primary groundwater evaluation model are calculated using the parameter optimization algorithm. The parameters of the primary groundwater evaluation model are optimized according to the optimal model parameters to obtain the groundwater evaluation model. By constructing a groundwater evaluation model to evaluate groundwater, the evaluation results can be accurate and the evaluation efficiency is higher, and continuous analysis of groundwater in the target area can be achieved. The pre-constructed water quality training set is groundwater information pre-marked with groundwater grades, including contaminated groundwater information, groundwater information with reduced water levels, and groundwater information that meets normal groundwater standards. Using the parameter optimization algorithm to optimize the parameters of the primary groundwater evaluation model can speed up the model training process and further optimize the primary groundwater evaluation model, so as to obtain a groundwater evaluation model that completes the training and optimization process, making the evaluation results of the groundwater evaluation model more accurate and more widely used. Finally, the core water quality information is input into the groundwater evaluation model to evaluate the groundwater grade and obtain the groundwater quality evaluation grade of the target area.

[0095] S108. Generate a groundwater detection report for the target area by combining groundwater level information, groundwater monitoring information and groundwater evaluation level.

[0096] In this embodiment, since the groundwater level information can reflect the global water level information of the target area, the monitoring point water level information in the groundwater monitoring information is the local water level information of the target area, and the monitoring point water level information is more accurate than the groundwater level information. Therefore, a data fusion algorithm is used to fuse the groundwater level information with the monitoring point water level information to obtain accurate global groundwater level information of the target area. The global groundwater level information, groundwater monitoring information and groundwater evaluation level are combined to generate a groundwater detection report for the target area. The groundwater detection report can display the groundwater distribution of all grid areas in the target area, as well as the groundwater level and groundwater evaluation level.

[0097] In one embodiment, the surface remote sensing information includes optical remote sensing information, microwave remote sensing information and infrared remote sensing information. The surface characteristic parameters of the target area are obtained based on the surface remote sensing information. After weights are assigned to the surface characteristic parameters according to meteorological information, the groundwater level information of the target area is calculated using an inversion algorithm, which includes the following steps:

[0098] Analyze the band reflectivity in optical remote sensing information and invert the surface vegetation index of the target area;

[0099] Extract the echo intensity of microwave signals in microwave remote sensing information, and calculate the surface soil moisture of the target area based on the echo intensity;

[0100] Obtain the surface temperature characteristics of the target area through infrared remote sensing information;

[0101] The surface vegetation index, surface soil moisture and surface temperature characteristics are integrated to obtain the surface characteristic parameters of the target area;

[0102] Acquire regional spatial temperature information and regional precipitation information of the target area according to meteorological information, and assign weights to surface characteristic parameters according to the regional spatial temperature information and regional precipitation information;

[0103] The groundwater level information of the target area is calculated based on the surface vegetation index, surface soil moisture and surface temperature with completed weight distribution and using the inversion algorithm.

[0104] In this embodiment, the surface remote sensing information includes optical remote sensing information, microwave remote sensing information and infrared remote sensing information. Optical remote sensing information is obtained by an optical remote sensing satellite equipped with an optical sensor by capturing and analyzing electromagnetic wave signals such as visible light and infrared light reflected, scattered and emitted by the surface. Microwave remote sensing information is the microwave radiation information of the earth's surface captured by the microwave remote sensing satellite through the microwave sensor. Microwave remote sensing has the ability to penetrate clouds and fog and can be carried out all day without being affected by weather. It uses the interaction between microwaves and surface materials to obtain information such as scattering, absorption and reflection of the surface. Infrared remote sensing information is the thermal radiation information of the earth's surface captured by the thermal infrared remote sensing satellite through the infrared sensor. The infrared remote sensing detection band is generally between 0.76-1000 microns. It uses infrared remote sensors to detect the information of the difference in infrared characteristics reflected or radiated by vegetation and other objects at a long distance, and can also reflect the temperature distribution of the surface.

[0105] The surface vegetation index refers to the normalized difference vegetation index (NDVI), which can reflect the growth status and biomass of vegetation. The surface vegetation index can be calculated based on the radiance values ​​of the infrared band and the red band in optical remote sensing information. The calculation formula is NDVI = (NIR-RED) / (NIR+RED), where NIR refers to the radiance value of the near-infrared band and Red refers to the radiance value of the red band. When the groundwater level is high, vegetation may be more likely to obtain water and grow more luxuriantly, otherwise its growth may be restricted. Of course, if the groundwater level is too high, it may also lead to soil salinization, which will seriously affect the growth and development of plants. Therefore, the groundwater level has a strong control effect on vegetation within a certain range. Vegetation is equivalent to an indicator of groundwater. Extracting the vegetation index through remote sensing means can indirectly invert the information of the groundwater level. There is a linear regression relationship between the groundwater level and groundwater.

[0106] Surface soil moisture refers to the water content of the soil. When the groundwater level is high, the surface soil can be replenished by capillary water to maintain a high soil moisture content. As the groundwater level decreases, the soil moisture content decreases as the capillary water replenishment decreases. Therefore, the groundwater level can be inverted based on the surface soil moisture. The steps to obtain surface soil moisture based on microwave remote sensing information are: first, perform atmospheric correction on the microwave remote sensing information to eliminate the influence of the atmosphere on the microwave remote sensing information, then calculate the dielectric constant of the soil based on the Dobson dielectric model (Dobson model), and finally calculate the soil moisture content based on the dielectric constant, that is, the surface soil moisture.

[0107] The inversion steps of the surface temperature characteristics of the target area include calibrating the infrared remote sensing information, converting it into radiance, using the atmospheric correction model (such as MODTRAN, SBDART) to correct the atmospheric influence, and then converting the surface radiance into surface temperature characteristics according to Planck's blackbody radiation law. The calculation formula is as follows:

[0108]

[0109] Among them, K1 and K2 are the calibration constants of the infrared sensor, L is the surface radiation brightness, and T is the surface temperature characteristic.

[0110] Since groundwater affects the temperature of the surface soil through heat conduction and evaporation heat absorption, the degree of groundwater enrichment and the height of the groundwater level directly affect the temperature characteristics of the soil. Therefore, the surface temperature characteristics detected by remote sensing technology can be used to invert the groundwater level.

[0111] Due to different climatic conditions, the reliability of surface vegetation index, surface soil moisture and surface temperature characteristics is different. For example, in humid areas with high atmospheric precipitation, precipitation will change the capillary water distribution of the soil profile. The water infiltrated by atmospheric precipitation can only be inverted after evaporation. Due to continuous precipitation, atmospheric precipitation cannot evaporate. Therefore, it is not accurate to invert the groundwater level through surface soil moisture. At this time, the weight of surface soil moisture needs to be reduced. Although the surface temperature characteristics can also be applied in humid areas with high atmospheric precipitation, the inversion accuracy will be reduced under low temperature conditions such as winter. This is because groundwater affects the temperature of the surface soil through heat conduction effect and evaporation heat absorption effect, and the heat source effect of groundwater in winter will interfere with the evaporative cooling effect and thus affect the inversion accuracy of the surface temperature characteristics. Therefore, it is necessary to assign weights to the surface characteristic parameters based on the regional spatial temperature information in the meteorological information (for example, the monthly average temperature in the target area is 24 degrees Celsius) and the regional precipitation information (for example, the monthly average precipitation in the target area is 114.3 mm). This method can ensure that no matter what the climatic conditions are, the three parameters of surface vegetation index, surface soil moisture and surface temperature characteristics can relatively accurately invert the groundwater level information of the target area, thereby making the subsequent distribution of groundwater monitoring points more reasonable.

[0112] After completing the weight distribution, the groundwater level information of the target area is inverted by combining the surface vegetation index, surface soil moisture and surface temperature characteristics. Specifically, the surface vegetation index, surface soil moisture and surface temperature characteristics are input into the groundwater inversion model pre-constructed according to the geological conditions and climatic conditions of the target area to obtain the groundwater level information of the target area. The groundwater inversion model is as follows:

[0113]

[0114] Among them, H is the groundwater level information, μ1, μ2, and μ3 are the weights of surface soil moisture A, surface vegetation index B, and surface temperature characteristics C, respectively, h is the penetration depth of the remote sensing band emitted by the microwave sensor, and A m A is the maximum surface soil moisture when groundwater approaches or overflows the surface. h is the surface soil moisture at the underground depth of h, A0 is the surface soil moisture on the soil surface, ∈, η, κ are fixed coefficients.

[0115] In one embodiment, the surface remote sensing information also includes a surface remote sensing image. Acquiring the surface water distribution characteristics of the target area according to the surface remote sensing information and generating a monitoring point distribution strategy for the target area in combination with the groundwater level information and the surface water distribution characteristics include the following steps:

[0116] The optical reflection characteristics of the target area are extracted through optical remote sensing information, and the surface water area of ​​the target area is marked according to the optical reflection characteristics;

[0117] Divide the target area into grids according to the surface remote sensing image to obtain multiple grid areas;

[0118] For each grid area of ​​the target region, the area of ​​the surface water region within the grid area is calculated to obtain the surface water area;

[0119] The number of monitoring points in each grid area is set based on the surface water area and the groundwater level information corresponding to the grid area, and the location of the monitoring points in the grid area is generated according to the number of monitoring points;

[0120] The locations of all monitoring points in all grid areas are integrated to obtain the monitoring point distribution strategy for the target area.

[0121] In this embodiment, the surface water area refers to the area where rivers and streams and other water bodies are distributed on the surface. When visible light, near-infrared light and other light are irradiated on the surface of the water body, reflection, scattering and absorption phenomena will occur. For example, the reflectivity of the blue-green band in the visible light band is higher than that of the red band, while the reflection energy of the near-infrared and short-wave infrared parts is less. This is significantly different from the spectral characteristics of vegetation, which helps to distinguish water bodies from land. After the surface water body is identified, it is marked in the surface remote sensing image. The specific calculation formula is NDWI = (G-NIR) / (G + NIR), where G is the reflectivity of the green band, NIR is the reflectivity of the near-infrared band, and NDWI refers to the normalized difference water index. The area with a normalized difference water index greater than zero is screened out to obtain the surface water area.

[0122] Then, the target area is gridded according to the surface remote sensing image of the target area (it can be divided into multiple grid areas with approximately equal areas) to obtain multiple grid areas. The area measurement tool in the geographic information system software or remote sensing processing software can be used to calculate the area of ​​the water body marked in each grid area in the surface remote sensing image to obtain the surface water area of ​​each grid area. In addition, the surface water area of ​​the grid area can also be calculated using the filling method or curve fitting method. Taking the filling method as an example, the irregular surface water area in the grid area is divided into several parts, and then these several parts are recombined to obtain a more regular surface water area, and then the corresponding area calculation formula is used to calculate the area of ​​the regularized surface water area to obtain the surface water area of ​​all grid areas.

[0123] The number of monitoring points in each grid area is set according to the groundwater level information corresponding to the surface water area and the grid area. For example, for grid areas with large surface water areas and deep groundwater levels, the number of monitoring points set is also more. Similarly, for grid areas with small surface water areas or no surface water areas and deep groundwater levels, the number of monitoring points is also less. According to the number of monitoring points, the corresponding number of groundwater monitoring points are set in each grid area, which can be evenly distributed in the grid area. Integrating the positions of all monitoring points in all grid areas is the monitoring point distribution strategy for the target area. Setting the monitoring point distribution strategy according to the groundwater level and surface water distribution can maximize the guarantee of comprehensive monitoring of groundwater in the target area, ensure that certain key areas are not missed, and ensure that the final monitoring results are more accurate. At the same time, for some unimportant areas, reducing the distribution of monitoring points can reduce unnecessary waste of resources. In addition, monitoring points are set according to the distribution of surface water because surface water is one of the important sources of water for human life, including agricultural irrigation, industrial water, domestic water, etc., while groundwater will replenish surface water through springs, seepage and injection, and is an important source for maintaining the water levels of rivers and lakes. Once the groundwater around surface water is polluted, it is likely to pollute the surface water. Therefore, it is necessary to focus on monitoring the groundwater in the areas around surface water.

[0124] In one embodiment, obtaining monitoring information of all groundwater monitoring stations through the Beidou communication unit and the Bluetooth communication unit, and integrating all monitoring information to obtain groundwater monitoring information of the target area includes the following steps:

[0125] Acquire the first groundwater information collected by all groundwater monitoring stations through the Beidou communication unit;

[0126] Matching the first groundwater information with the corresponding monitoring point locations, and marking the monitoring point locations where information is missing as abnormal monitoring points;

[0127] For each grid area, calculating the information difference between all first groundwater information in the grid area;

[0128] Screen out the first groundwater information whose information difference is greater than a preset difference threshold, and mark the monitoring point position corresponding to the first groundwater information as an abnormal monitoring point;

[0129] According to the monitoring point location corresponding to the abnormal monitoring point and using the shortest path algorithm, an information collection route for the monitoring drone is generated, and the monitoring drone is equipped with a Bluetooth communication module;

[0130] Control the monitoring drone to obtain the second groundwater information of all abnormal monitoring points through the Bluetooth communication unit according to the information collection route;

[0131] The first groundwater information that has completed information screening is integrated with the second groundwater to obtain the groundwater monitoring information of the target area.

[0132] In this embodiment, the first groundwater information collected by the groundwater monitoring station is first transmitted to the groundwater monitoring platform through the Beidou communication unit, and the received first groundwater information is matched with the corresponding monitoring point location, that is, the source location of all the first groundwater information is found, and then the information integrity of all the first groundwater information is identified to analyze whether there is information vacancy. For example, there is no pH value in a certain first groundwater information, indicating that the first groundwater information has information vacancy. When there is information vacancy, the monitoring point location corresponding to the groundwater monitoring information with information vacancy is marked as an abnormal monitoring point. Then, the information difference between the first groundwater information in each grid area is calculated, and the first groundwater information with information difference greater than the preset difference threshold is screened out, and the corresponding monitoring point location is marked as an abnormal monitoring point. This is because the degree of data proximity between two groundwater monitoring points is related to the distance between them. The closer the distance between the two groundwater monitoring points, the closer the corresponding groundwater monitoring data. Therefore, if the information difference between the first groundwater information in the same grid area is large, it means that there is an abnormal value in the first groundwater information. The monitoring point position corresponding to the abnormal value is marked as an abnormal monitoring point, which is convenient for the subsequent use of the Bluetooth communication unit to re-acquire the monitoring data.

[0133] After marking all abnormal monitoring points, refer to Figure 2 According to the location of the abnormal monitoring point, the shortest path algorithm is used to generate the information collection route of the monitoring drone, and the monitoring drone is controlled to go to the abnormal monitoring point according to the information collection route, and the Bluetooth communication module installed on it is used to obtain the second groundwater information transmitted by the groundwater monitoring station through the Bluetooth communication unit. Commonly used shortest path optimization algorithms include Dijkstra algorithm, Bellman-Ford algorithm and bidirectional search algorithm. Taking the bidirectional search algorithm as an example, the search is performed from the starting point and the end point at the same time. When the two searches meet, the shortest path is found. The shortest path is generated for the monitoring drone through the shortest path algorithm, which can improve the information collection efficiency of the monitoring drone. The Bluetooth communication unit can realize the safe transmission of groundwater monitoring information at a very low cost, which can make up for the defects of the Beidou communication unit and complete the safe transmission of groundwater monitoring information together with the Beidou communication unit. After removing the abnormal information in the first groundwater information, it is integrated with the second groundwater information to obtain the complete monitoring information of the target area, that is, the groundwater monitoring information. By obtaining groundwater monitoring information twice, abnormal data caused by abnormalities in groundwater monitoring equipment or abnormal transmission of communication equipment can be eliminated to the greatest extent, ensuring that the groundwater monitoring information is more accurate and the subsequent groundwater evaluation results based on the groundwater monitoring information are more reliable.

[0134] In one embodiment, a groundwater evaluation model is constructed based on a parameter optimization algorithm and a neural network model, and groundwater monitoring information is input into the trained groundwater evaluation model to obtain a groundwater evaluation grade, including the following steps:

[0135] Wavelet transform is used to eliminate abnormal information in groundwater monitoring information, and the information gain of each water quality attribute information in the groundwater monitoring information after wavelet transform is calculated, and the core water quality information in the water quality attribute information is screened out according to the information gain;

[0136] A primary groundwater assessment model was constructed based on the convolutional neural network model and the long short-term memory network model;

[0137] The primary groundwater assessment model was trained using a pre-built water quality training set;

[0138] During the model training process, the optimal model parameters of the primary groundwater assessment model were calculated using a parameter optimization algorithm;

[0139] Optimize the parameters of the primary groundwater assessment model according to the optimal model parameters to obtain the groundwater assessment model;

[0140] The core water quality information is input into the trained groundwater assessment model to obtain the groundwater assessment grade.

[0141] In this embodiment, wavelet transform is used to remove abnormal information in groundwater monitoring information again, and then the information gain of each water quality attribute information in groundwater monitoring information is calculated, and the core water quality information in water quality attribute information is screened out according to the information gain. By this method, the abnormal change data of groundwater monitoring information in time series can be removed, for example, at a certain time node, the temperature data in groundwater monitoring information suddenly surges, and then suddenly decreases at the next time node. By calculating the information gain of each water quality attribute information in groundwater monitoring information, it is possible to screen out a number of water quality attribute information that has a greater impact on groundwater quality, thereby simplifying groundwater monitoring information, improving the efficiency of subsequent groundwater evaluation, and reducing the amount of calculation. Real-time groundwater monitoring and rapid analysis of groundwater monitoring information can achieve real-time grasp of groundwater water quality and water level changes. Once an emergency such as water level drop and water quality deterioration occurs, it can be discovered in time and countermeasures can be quickly formulated. For example, when groundwater is polluted, the pollutants and pollution sources in the groundwater can be traced immediately, providing a scientific basis for pollution prevention and control, and preventing pollution from spreading.

[0142] A primary groundwater evaluation model is constructed based on the convolutional neural network model and the long short-term memory network model. The primary groundwater evaluation model is trained using the pre-constructed water quality training set. At the same time, the optimal model parameters of the primary groundwater evaluation model are calculated using the parameter optimization algorithm. The parameters of the primary groundwater evaluation model are optimized according to the optimal model parameters to obtain the groundwater evaluation model. By constructing a groundwater evaluation model to evaluate groundwater, the evaluation results can be accurate and the evaluation efficiency is higher. It can realize continuous analysis of groundwater in the target area. The primary groundwater evaluation model is constructed based on the convolutional neural network model and the long short-term memory network model. It can not only extract the spatial characteristics of groundwater monitoring information at different monitoring points in the spatial dimension, but also capture the changing trend of groundwater quality and water level over time, so as to more comprehensively evaluate the groundwater situation. The pre-constructed water quality training set is groundwater information pre-marked with groundwater grades. These groundwater information include polluted groundwater information, groundwater information with reduced water level, and groundwater information that meets normal groundwater standards. Using the parameter optimization algorithm to optimize the parameters of the primary groundwater evaluation model can speed up the model training process and further optimize the primary groundwater evaluation model, so as to obtain a groundwater evaluation model that has completed the training and optimization process, making the evaluation results of the groundwater evaluation model more accurate and more widely used. Finally, the core water quality information is input into the groundwater evaluation model to evaluate the groundwater level and obtain the groundwater quality evaluation level of the target area.

[0143] In one embodiment, wavelet transform is used to remove abnormal information in groundwater monitoring information, and the importance of water quality attributes of various water quality attribute information in the groundwater monitoring information after wavelet transform is calculated, and the core water quality information in the water quality attribute information is screened out according to the importance of water quality attributes, including the following steps:

[0144] The groundwater monitoring information is decomposed at multiple scales using wavelet basis functions to obtain multiple groundwater monitoring information components and high-frequency coefficients corresponding to the groundwater monitoring information components;

[0145] Eliminate groundwater monitoring information components whose high-frequency coefficients are less than a preset coefficient threshold in groundwater monitoring information components;

[0146] Reconstruct the removed groundwater monitoring information components to obtain normal groundwater monitoring information;

[0147] Calculate the information entropy of each water quality attribute information in normal groundwater monitoring information, and calculate the information gain of the corresponding water quality attribute information based on the information entropy;

[0148] All water quality attribute information whose information gain is greater than a preset gain threshold is screened out as core water quality information.

[0149] In this embodiment, wavelet transform is a digital signal processing technology, which converts a large range of data into a series of feature quantities through operations such as decomposition, scaling and translation. The basic principle of using wavelet transform for denoising is to decompose the original signal into several wavelet wavelets of specific frequencies, and then obtain the corresponding wavelet coefficients at various frequencies. For a noisy signal, its high-frequency coefficients are often dominated by noise, while the low-frequency coefficients often correspond to the basic information of the signal. Therefore, the wavelet denoising method is to remove the noise of the high-frequency signal while keeping the low-frequency signal unchanged.

[0150] Select appropriate wavelet basis (such as Daubechies, Haar, Symlet, etc.) and decomposition layers to perform wavelet decomposition on groundwater monitoring information to obtain multiple groundwater monitoring information components of different frequencies and high-frequency coefficients corresponding to the groundwater monitoring information components. After eliminating the groundwater monitoring information components whose high-frequency coefficients are less than the preset coefficient threshold, reconstruct the eliminated groundwater monitoring information components to obtain normal groundwater monitoring information. The wavelet transform reconstruction principle is to recombine the decomposed wavelet coefficients into the original signal through the inverse process of wavelet transform to obtain complete normal groundwater monitoring information with almost no abnormal data. After completing the wavelet transform, calculate the overall entropy of each water quality attribute information in the normal groundwater monitoring information. The formula for calculating the overall entropy is as follows:

[0151]

[0152] Among them, P(x i ) is the probability of occurrence of the i-th water quality attribute information in the groundwater monitoring information, X is the total set of all water quality attribute information, and n is the number of water quality attribute information.

[0153] After obtaining the overall entropy of each water quality attribute information, the conditional entropy of all water quality attribute information is calculated respectively. The formula for calculating the conditional entropy is as follows:

[0154]

[0155] Where D is the data set of all data of water quality attribute information A, v is the number of data in data set D, and D j is the jth data in the data set, H(D j ) is D j The overall entropy of .

[0156] After calculating the overall entropy and conditional entropy, the conditional entropy is subtracted from the overall entropy to obtain the information gain of the water quality attribute information. After calculating the information gain of all water quality attribute information, all water quality attribute information with information gain greater than the preset gain threshold is screened out as the core water quality information. Information gain is a concept in information theory that measures the degree of uncertainty reduction of a feature under a given category. By calculating the information gain, the most representative features can be selected, thereby improving the prediction performance of the model, reducing the amount of calculation of the model, and improving the efficiency of groundwater information evaluation.

[0157] In one embodiment, calculating the optimal model parameters of the primary groundwater assessment model using a parameter optimization algorithm includes the following steps:

[0158] The initial population of the parameter optimization algorithm is constructed using multiple parameters to be optimized of the primary groundwater assessment model;

[0159] Initializing an initial population, and inputting the initialized initial population into a primary groundwater assessment model to obtain an initial output value of the primary groundwater assessment model;

[0160] Calculate the root mean square error between the initial output value and the pre-acquired benchmark output value, and use the root mean square error as the fitness of the initial population;

[0161] The individuals in the initial population whose fitness is greater than the preset first fitness threshold are retained to obtain the core population;

[0162] The population individuals whose fitness in the core population is greater than a preset second fitness threshold are taken as core individuals, and the second fitness threshold is greater than the first fitness threshold;

[0163] Based on the core individuals and using the dynamic weight coefficient strategy, the core population is updated to the target core population;

[0164] Use the Levy flight strategy to update and adjust the core individuals with the best fitness in the target core population, and calculate the updated target fitness of all population individuals in the target core population;

[0165] The population updating iteration steps are repeated for the target core population until the preset maximum number of iterations is reached, and the optimal model parameters of the primary groundwater assessment model are obtained.

[0166] In this embodiment, refer to Figure 3, the parameter optimization algorithm is an optimized and improved gray wolf optimization algorithm. The gray wolf optimization algorithm is a group intelligence optimization algorithm that simulates the hierarchy and hunting behavior of wolves. It has the advantages of few adjustable parameters, intuitive and simple principles, and strong global search capabilities. However, in the traditional gray wolf optimization algorithm, all gray wolf individuals move towards wolf α. If wolf α is not the representative of the global optimal solution, it will reduce the diversity of the population and make the algorithm easy to fall into the local optimal solution. Therefore, the Levy flight strategy is introduced to expand the search domain, increase the diversity of solutions, and thus enhance the algorithm's exploration ability in the global range. The Levy flight strategy is a random behavior strategy used to simulate the step size and direction in the random walk or search process. It can simulate the movement of organisms when searching for food or resources in some cases. At the same time, in order to avoid converging to the local solution too early, the dynamic weight coefficient is used to update the position of the wolf pack. In addition, in order to further optimize the gray wolf optimization algorithm, the convergence factor in the gray wolf optimization algorithm is replaced with a nonlinear convergence factor. Using the nonlinear convergence factor to iterate the population can prompt the population to search extensively in the search space, which is conducive to avoiding the population from falling into the local optimal value, and in the later stage of iteration, it can accelerate the convergence of the population through fine search in a small range.

[0167] The initial population of the parameter optimization algorithm is constructed from multiple parameters to be optimized of the primary groundwater evaluation model. The parameters to be optimized include the initial learning rate, regularization coefficient, and number of neurons in the hidden layer of the primary groundwater evaluation model. The initial population is initialized for the purpose of introducing diverse individuals, constructing constraints, defining population size, and setting the maximum number of iterations. The initialized initial population is input into the primary groundwater evaluation model for training iteration to obtain the initial output value of the primary groundwater evaluation model, and the root mean square error between the initial output value and the pre-acquired benchmark output value is calculated according to the root mean square error formula, and the root mean square error is used as the fitness of the initial population. Since the main purpose of this embodiment is to optimize the parameters of the primary groundwater evaluation model, the root mean square error of the primary groundwater evaluation model is used as the fitness of the initial population. The individuals in the initial population whose fitness is greater than the preset first fitness threshold are retained to obtain the core population, and the individuals in the core population whose fitness is greater than the preset second fitness threshold are taken as core individuals. Since the role of the core individual is to guide other individuals in the core population to update and iterate, the core individual must be greater than other individuals, that is, the second fitness threshold is greater than the first fitness threshold.

[0168] According to the position of the core individuals and the dynamic weight coefficient strategy, the core population is dynamically updated to obtain the target core population. In each dynamic update process, the weight coefficient is dynamically adjusted according to the current number of iterations and the fitness of the population to prevent the algorithm from converging to the local optimal solution too early. After completing the update of the core population, the core individuals with the best fitness in the target core population are updated and adjusted using the Levy flight strategy, and their fitness is recalculated. If the fitness decreases, it is retained in the target core population. If the fitness increases, the core individuals with the best fitness after the update are removed from the target core population to prevent the parameter optimization algorithm from falling into the local optimal solution. Then the convergence factor is updated and adjusted using the nonlinear adjustment strategy to obtain the nonlinear convergence factor. The subsequent iteration process uses the nonlinear convergence factor to adjust the step size or moving distance when the gray wolf individual updates its position. The updated target fitness of all population individuals in the target core population is calculated, and the population update iteration steps are repeated for the target core population until the preset maximum number of iterations is reached. The individuals with the highest fitness in the target core population after multiple iterations are decoded as the optimal model parameters of the primary groundwater assessment model.

[0169] In one embodiment, the core water quality information is input into the trained groundwater evaluation model to obtain the groundwater evaluation level of the target area, including the following steps:

[0170] The core water quality information is input into the spatial feature extraction module of the groundwater assessment model to extract spatial features and obtain the local groundwater characteristics of the target area. The spatial feature extraction module is constructed based on the convolutional neural network model.

[0171] The core water quality information is input into the time series feature extraction module of the groundwater assessment model for time series feature analysis to obtain the global groundwater characteristics of the target area within a preset time period. The time series feature extraction module is constructed based on the long short-term memory neural network model.

[0172] The local groundwater characteristics and global groundwater characteristics are input into the multi-scale feature fusion module of the groundwater assessment model for feature fusion to obtain the groundwater fusion characteristics of the target area;

[0173] After the groundwater fusion features are input into the fully connected layer of the groundwater assessment model, the groundwater level in the target area is evaluated using the pre-set classification function to obtain the groundwater evaluation level of the target area.

[0174] In this embodiment, the core water quality information is input into the spatial feature extraction module of the groundwater evaluation model for spatial feature extraction. The spatial feature extraction module can be composed of multiple convolutional layers, activation layers, pooling layers, and fully connected layers. The convolutional layer uses the convolution kernel to extract local features at a certain step size, extracts features of different time scales through sliding window operations, and then introduces nonlinear features in the activation layer to avoid gradient disappearance. In the pooling layer, the dimension is reduced by maximum pooling or average pooling, and the main features are retained to reduce the amount of calculation when the subsequent model performs grade evaluation. The fully connected layer further maps the features of the pooling layer, and finally outputs the local groundwater features of the target area. The local groundwater features represent the groundwater level features and water quality features at different locations in the target area. The core idea of ​​the convolutional neural network is to extract features through operations such as convolution and pooling, and then map the input data to a high-dimensional feature space. Finally, the features are classified or regressed through the fully connected layer. The convolutional neural network model includes convolutional layers, activation layers, pooling layers, and fully connected layers. The core water quality information is input into the time series feature extraction module of the groundwater assessment model for time series feature analysis. The time series feature extraction module may include multiple LSTM layers (e.g., 3 layers). The flow of information is controlled by the input gate, forget gate, and output gate in the time series feature extraction module. The input gate processes the current input and the hidden state of the previous moment to determine which information is important and allows this information to update the state of the memory unit. The forget gate evaluates which information is no longer important based on the current input and the hidden state of the previous moment, and then deletes this information from the memory unit to maintain the relevance of the memory. The output gate determines the output content based on the current unit state and input, which in turn affects the hidden state of the next moment and subsequent calculations. Through the synergistic effect of these three gates, the changing characteristics of groundwater in the target area within a certain period of time, that is, the global groundwater characteristics, can be effectively captured.

[0175] The local groundwater characteristics and the global groundwater characteristics are input into the multi-scale feature fusion module of the groundwater evaluation model for feature fusion. The multi-scale feature fusion module can use the attention mechanism or the gating mechanism to effectively fuse the spatial features and the temporal features. Taking the attention mechanism as an example, the correlation score between the two features can be calculated, and then weighted fusion can be performed according to the score. The groundwater fusion features obtained by feature fusion can describe the spatial distribution characteristics and temporal variation characteristics of the groundwater information in the target area. After the groundwater fusion features are input into the fully connected layer of the groundwater evaluation model, the groundwater in the target area is graded using the pre-set classification function. The pre-set classification function can be the Softmax function (normalized exponential function), which can grade the groundwater fusion features and divide different groundwater fusion features into different groundwater evaluation grades.

[0176] In one embodiment, generating a groundwater detection report for a target area by combining groundwater level information, groundwater monitoring information, and groundwater evaluation level includes the following steps:

[0177] Extracting water level information of monitoring points from groundwater monitoring information;

[0178] The water level information of the monitoring points is integrated with the groundwater level information to obtain the global groundwater level information of the target area;

[0179] A groundwater detection report for the target area is generated by combining the global groundwater level information, groundwater monitoring information and groundwater evaluation level.

[0180] In this embodiment, combining groundwater level information, groundwater monitoring information and groundwater evaluation grade to generate a groundwater detection report for the target area is a comprehensive process, and a multi-dimensional analysis is required to comprehensively evaluate the groundwater resource status. The core of this step is to deeply integrate and analyze these three types of key data to provide comprehensive and accurate groundwater resource evaluation results. Groundwater level information usually includes water level height data at different monitoring points at different times. Groundwater monitoring information may include water quality parameters (such as pH value, total dissolved solids, nitrate content, etc.) and water volume parameters (such as flow rate, flow, etc.). The groundwater evaluation grade is a comprehensive rating result obtained by the evaluation model based on the above parameters. In the preprocessing stage, these data need to be unified, including alignment of time scales (such as unification to daily or monthly scales), consistency of spatial reference systems, identification and processing of outliers, etc. This step ensures the data quality and comparability of subsequent analysis.

[0181] Next, spatial analysis is performed. The Kriging interpolation method is used to transform the discrete monitoring point data into a continuous spatial distribution surface. The Kriging method takes into account spatial autocorrelation and can provide the best linear unbiased estimate, which is particularly suitable for natural phenomena with spatial continuity such as groundwater. Spatial distribution maps are generated for groundwater levels, main water quality parameters (such as total dissolved solids content) and evaluation levels. These spatial distribution maps not only intuitively show the spatial variation characteristics of each parameter, but also provide a basis for identifying potential problem areas. For example, the water level distribution map can reveal over-exploitation areas or recharge areas; the water quality parameter distribution map can indicate the location of potential pollution sources; and the evaluation level distribution map comprehensively reflects the spatial differences in the status of groundwater resources. Through the overlay analysis of these spatial distribution maps, areas that need to be focused on can be identified, such as areas with significant water level decline and poor water quality.

[0182] Time series analysis is the next key step. The water level and main water quality parameter data of each monitoring point are decomposed into trend terms, seasonal terms and residual terms. This decomposition helps to understand the long-term change trend, periodic fluctuations and short-term fluctuation characteristics of the groundwater system. The Mann-Kendall trend test is used for the trend term to quantitatively evaluate the significance of the change trend. Fourier analysis is performed on the seasonal term to identify the main periodic characteristics, which helps to understand the response pattern of the groundwater system to climate change or human activities. Combining the results of spatial analysis and time series analysis, a dynamic change map of the groundwater system can be drawn. For example, a spatial distribution map of the interannual change rate of the water level can be generated to intuitively show the spatial differences in water level changes in different regions; or a spatiotemporal distribution map of the frequency of exceeding the standard of water quality parameters can be generated to reflect the spatiotemporal evolution characteristics of water quality problems.

[0183] Correlation analysis is an important means of integrating different types of data. The partial correlation analysis method is used to analyze the relationship between groundwater level, main water quality parameters and evaluation level while controlling the influence of other factors. This analysis can reveal some important associations, such as whether the decline in water level is significantly correlated with the deterioration of certain water quality parameters, or whether the change in evaluation level is mainly affected by water level or water quality. The quantification of these relationships helps to understand the dynamic mechanism of the groundwater system and provide a basis for management decisions.

[0184] Based on the above analysis, a groundwater condition prediction model was constructed using the random forest algorithm. The random forest model can handle nonlinear relationships, is insensitive to outliers, and can provide feature importance assessments. Using historical water levels, water quality parameters, and related environmental factors (such as rainfall, land use, etc.) as inputs, the groundwater conditions in a certain period of time in the future are predicted. This prediction includes not only changes in water levels and major water quality parameters, but also potential changes in the comprehensive evaluation level. By setting different scenarios (such as extreme climate events, land use changes, etc.), scenario analysis is conducted to provide decision makers with possible results under different conditions. Finally, the results of the multidimensional analysis are integrated into the groundwater detection report. The structure of the report should include:

[0185] 1. Executive Summary: Summarizes the main findings and recommendations.

[0186] 2. Research background: Introduce the geographical, geological and hydrological characteristics of the target area.

[0187] 3. Data and methods: Detailed description of data sources, pre-processing methods and analysis techniques.

[0188] 4. Result analysis:

[0189] a) Groundwater level status: including the spatial distribution characteristics of the water level, temporal variation trends and abnormal area analysis.

[0190] b) Water quality status: spatial distribution, temporal evolution and exceedance analysis of major water quality parameters.

[0191] c) Evaluation level analysis: spatial distribution and temporal changes of different levels and their relationship with water level and water quality.

[0192] d) Correlation analysis: the relationship between water level, water quality and evaluation level and their environmental influencing factors.

[0193] e) Forecast results: Forecast of future groundwater conditions and analysis under different scenarios.

[0194] 5. Discussion: Explain the observed phenomena and analyze possible causes and potential impacts.

[0195] 6. Conclusions and Recommendations: Summarize the main findings and put forward targeted management recommendations and future monitoring plans.

[0196] 7. Appendix: Contains detailed data tables, analysis charts and technical details.

[0197] The report should use a large number of visual charts, such as groundwater level contour maps, water quality parameter spatial distribution heat maps, time series trend maps, correlation matrix heat maps, etc., to intuitively display the analysis results. At the same time, the report should clearly explain the uncertainty and limitations of each analysis to ensure that decision makers can accurately understand and use this information. Through this multi-dimensional analysis and reporting method, the current status, changing trends and potential problems of groundwater resources in the target area can be fully reflected, providing scientific and reliable decision-making support for water resources management and environmental protection. The advantage of this method is that it integrates information on three key aspects: water level, water quality and evaluation level, provides a comprehensive view of the groundwater system, and helps to formulate more accurate and effective groundwater resource management strategies.

[0198] The present application also discloses a groundwater monitoring system based on Beidou dual-channel and Bluetooth communication, which is characterized by comprising:

[0199] a memory configured to store instructions; and

[0200] A processor is configured to call instructions from a memory and implement a method for groundwater monitoring based on Beidou dual-channel and Bluetooth communication according to any of the above items when executing the instructions.

[0201] Among them, the processor can adopt a central processing unit (CPU). Of course, according to actual usage, other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. can also be adopted. The general-purpose processor can adopt a microprocessor or any conventional processor, etc., and this application does not impose any restrictions on this.

[0202] Among them, the memory can be an internal storage unit of a computer device, such as a hard disk or memory of a computer device, or an external storage device of a computer device, such as a plug-in hard disk, a smart memory card (SMC), a secure digital card (SD) or a flash memory card (FC) equipped on the computer device, etc., and the memory can also be a combination of an internal storage unit and an external storage device of a computer device. The memory is used to store computer programs and other programs and data required by the computer device. The memory can also be used to temporarily store data that has been output or is to be output, and this application does not impose any restrictions on this.

[0203] An embodiment of the present application also provides a machine-readable storage medium, on which instructions are stored, and the instructions are used to enable a machine to execute the above-mentioned method of groundwater monitoring based on Beidou dual-channel and Bluetooth communication.

[0204] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0205] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0206] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0207] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0208] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0209] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0210] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0211] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0212] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A groundwater monitoring method based on Beidou dual-channel and Bluetooth communication, characterized in that: The method comprises the following steps: Obtain surface remote sensing information and meteorological information of the target area through satellite technology; Acquire the surface characteristic parameters of the target area based on the surface remote sensing information, assign weights to the surface characteristic parameters according to the meteorological information, and calculate the groundwater level information of the target area using an inversion algorithm; Acquire the surface water area of ​​the target area according to the surface remote sensing information, and generate a monitoring point distribution strategy for the target area in combination with the groundwater level information and the surface water area; Deploy a plurality of groundwater monitoring stations in the target area according to the monitoring point distribution strategy, wherein the groundwater monitoring stations are provided with an information transmission module, and the information transmission module includes a Beidou communication unit and a Bluetooth communication unit; Using the groundwater monitoring station to monitor groundwater in the target area; Acquire monitoring information of all the groundwater monitoring stations through the Beidou communication unit and the Bluetooth communication unit, and integrate all the monitoring information to obtain groundwater monitoring information of the target area; Constructing a groundwater evaluation model based on a parameter optimization algorithm and a neural network model, inputting the groundwater monitoring information into the trained groundwater evaluation model, and obtaining the groundwater evaluation grade of the target area; A groundwater detection report for the target area is generated by combining the groundwater level information, the groundwater monitoring information and the groundwater evaluation grade.

2. The method according to claim 1, characterized in that: The surface remote sensing information includes optical remote sensing information, microwave remote sensing information and infrared remote sensing information. The step of acquiring the surface characteristic parameters of the target area based on the surface remote sensing information, assigning weights to the surface characteristic parameters according to the meteorological information and calculating the groundwater level information of the target area using an inversion algorithm includes the following steps: Analyze the band reflectance in the optical remote sensing information to invert the surface vegetation index of the target area; Extracting the echo intensity of the microwave signal in the microwave remote sensing information, and calculating the surface soil moisture of the target area according to the echo intensity; Acquiring the surface temperature characteristics of the target area through the infrared remote sensing information; Integrating the surface vegetation index, the surface soil moisture and the surface temperature characteristics to obtain the surface characteristic parameters of the target area; Acquiring regional spatial temperature information and regional precipitation information of the target area according to the meteorological information, and assigning weights to the surface characteristic parameters according to the regional spatial temperature information and the regional precipitation information; The groundwater level information of the target area is calculated based on the surface vegetation index, the surface soil moisture and the surface temperature after weight distribution and using an inversion algorithm.

3. The method according to claim 2, characterized in that The surface remote sensing information also includes a surface remote sensing image. The step of acquiring the surface water distribution characteristics of the target area according to the surface remote sensing information and generating a monitoring point distribution strategy for the target area in combination with the groundwater level information and the surface water distribution characteristics includes the following steps: Extracting optical reflection characteristics of the target area through the optical remote sensing information, and marking the surface water area of ​​the target area according to the optical reflection characteristics; Dividing the target area into grids according to the surface remote sensing image to obtain a plurality of grid areas; For each of the grid areas of the target area, calculating the area of ​​the surface water region within the grid area to obtain the surface water area; The number of monitoring points in each grid area is set in combination with the surface water area and the groundwater level information corresponding to the grid area, and the location of the monitoring points in the grid area is generated according to the number of monitoring points; The monitoring point locations of all the grid areas are integrated to obtain a monitoring point distribution strategy for the target area.

4. The method according to claim 3, characterized in that The step of acquiring the monitoring information of all the groundwater monitoring stations through the Beidou communication unit and the Bluetooth communication unit, and integrating all the monitoring information to obtain the groundwater monitoring information of the target area comprises the following steps: Acquire the first groundwater information collected by all the groundwater monitoring stations through the Beidou communication unit; Matching the first groundwater information with the corresponding monitoring point locations, and marking the monitoring point locations where information is missing as abnormal monitoring points; For each of the grid areas, calculating the information difference between all the first groundwater information within the grid area; Screening out the first groundwater information whose information difference is greater than a preset difference threshold, and marking the monitoring point position corresponding to the first groundwater information as the abnormal monitoring point; According to the monitoring point position corresponding to the abnormal monitoring point and using the shortest path algorithm, an information collection route of a monitoring drone is generated, wherein the monitoring drone is equipped with a Bluetooth communication module; Control the monitoring drone to obtain the second groundwater information of all the abnormal monitoring points through the Bluetooth communication unit according to the information collection route; The first groundwater information after information screening is integrated with the second groundwater to obtain the groundwater monitoring information of the target area.

5. The method according to claim 1, characterized in that The method of constructing a groundwater evaluation model based on a parameter optimization algorithm and a neural network model, inputting the groundwater monitoring information into the trained groundwater evaluation model, and obtaining a groundwater evaluation grade comprises the following steps: Using wavelet transform to eliminate abnormal information in the groundwater monitoring information, and calculating the information gain of each water quality attribute information in the groundwater monitoring information after the wavelet transform, and screening out the core water quality information in the water quality attribute information according to the information gain; A primary groundwater assessment model was constructed based on the convolutional neural network model and the long short-term memory network model; Using a pre-constructed water quality training set to perform model training on the primary groundwater assessment model; During the model training process, the optimal model parameters of the primary groundwater assessment model are calculated using a parameter optimization algorithm; Optimizing the parameters of the primary groundwater assessment model according to the optimal model parameters to obtain a groundwater assessment model; The core water quality information is input into the trained groundwater evaluation model to obtain a groundwater evaluation grade.

6. The method according to claim 5, characterized in that The method of eliminating abnormal information in the groundwater monitoring information by wavelet transform, calculating the importance of water quality attributes of each water quality attribute information in the groundwater monitoring information after wavelet transform, and selecting core water quality information in the water quality attribute information according to the importance of the water quality attributes comprises the following steps: Using wavelet basis functions to perform multi-scale decomposition on the groundwater monitoring information to obtain multiple groundwater monitoring information components and high-frequency coefficients corresponding to the groundwater monitoring information components; Eliminate the groundwater monitoring information components whose high-frequency coefficients are less than a preset coefficient threshold in the groundwater monitoring information components; Reconstructing the groundwater monitoring information components that have been eliminated to obtain normal groundwater monitoring information; Calculating the information entropy of each item of water quality attribute information in the normal groundwater monitoring information, and calculating the information gain of the corresponding water quality attribute information according to the information entropy; All the water quality attribute information whose information gain is greater than a preset gain threshold in the water quality attribute information are screened out as core water quality information.

7. The method according to claim 5, characterized in that The method of calculating the optimal model parameters of the primary groundwater assessment model using a parameter optimization algorithm comprises the following steps: constructing an initial population of a parameter optimization algorithm using a plurality of parameters to be optimized of the primary groundwater assessment model; Initializing the initial population, and inputting the initialized initial population into the primary groundwater evaluation model to obtain an initial output value of the primary groundwater evaluation model; Calculating a root mean square error between the initial output value and a pre-acquired reference output value, and using the root mean square error as the fitness of the initial population; Retaining the population individuals whose fitness in the initial population is greater than a preset first fitness threshold to obtain a core population; Taking the population individuals whose fitness in the core population is greater than a preset second fitness threshold as core individuals, where the second fitness threshold is greater than the first fitness threshold; Based on the core individuals and using a dynamic weight coefficient strategy, the core population is updated to a target core population; Using the Levy flight strategy to update and adjust the core individual with the best fitness in the target core population, and calculating the updated target fitness of all population individuals in the target core population; The population updating iteration step is repeated for the target core population until a preset maximum number of iterations is reached, thereby obtaining the optimal model parameters of the primary groundwater assessment model.

8. The method according to claim 5, characterized in that The step of inputting the core water quality information into the trained groundwater evaluation model to obtain the groundwater evaluation grade of the target area comprises the following steps: Inputting the core water quality information into the spatial feature extraction module of the groundwater assessment model to extract spatial features, thereby obtaining local groundwater features of the target area, wherein the spatial feature extraction module is constructed based on a convolutional neural network model; Inputting the core water quality information into the time series feature extraction module of the groundwater assessment model for time series feature analysis to obtain the global groundwater characteristics of the target area within a preset time period, wherein the time series feature extraction module is constructed based on a long short-term memory neural network model; Inputting the local groundwater characteristics and the global groundwater characteristics into the multi-scale feature fusion module of the groundwater assessment model for feature fusion to obtain the groundwater fusion characteristics of the target area; After the groundwater fusion features are input into the fully connected layer of the groundwater evaluation model, the groundwater level of the target area is evaluated using a preset classification function to obtain the groundwater evaluation level of the target area.

9. The method according to claim 1, characterized in that: The step of generating the groundwater detection report of the target area by combining the groundwater level information, the groundwater monitoring information and the groundwater evaluation grade comprises the following steps: Extracting water level information of monitoring points from the groundwater monitoring information; Fusing the monitoring point water level information with the groundwater level information to obtain global groundwater level information of the target area; A groundwater detection report for the target area is generated by combining the global groundwater level information, groundwater monitoring information and the groundwater evaluation grade.

10. A groundwater monitoring system based on Beidou dual-channel and Bluetooth communication, characterized in that: include: a memory configured to store instructions; as well as A processor is configured to call the instructions from the memory and to implement a method for groundwater monitoring based on Beidou dual-channel and Bluetooth communication according to any one of claims 1 to 9 when executing the instructions.

Images