Unmanned aerial vehicle monitoring system and method based on multi-area water sample collection and multi-sensor fusion

Abstract

The invention discloses an unmanned aerial vehicle monitoring system and method based on multi-region water sample collection and multi-sensor fusion. The unmanned aerial vehicle monitoring system comprises a telescopic sampling arm, a multi-sensor module, a data transmission module and an intelligent data processing module; the telescopic sampling arm is used for automatically collecting water samples in different water area depths; the multi-sensor module is used for monitoring a water sample in real time, accurately measuring multi-sensor data of a sampling point and transmitting the multi-sensor data to the intelligent data processing module in real time; the data transmission module is used for transmitting the acquired multi-sensor data to the intelligent data processing module in real time; and the intelligent data processing module is used for performing data fusion processing on the acquired multi-sensor data to generate an analysis report. Unique advantages of multi-region collection and data fusion are fully utilized, and comprehensive multi-point water sample collection and fusion processing of multi-sensor data in a deep water region are realized, so that the comprehensiveness of the sampled data and the accuracy of data analysis are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of UAV water sample collection and data processing, and specifically provides a UAV monitoring system and method based on multi-region water sample collection and multi-sensor fusion. Background Art

[0002] In recent years, UAV technology has been widely used in fields such as environmental monitoring, agriculture, and logistics. In terms of water sample collection and environmental monitoring, UAVs can reach areas that are difficult for humans to access, providing efficient and real-time data collection and analysis capabilities. However, there are still technical challenges in water sample collection and data processing in deep water areas, such as collection accuracy, sensor data fusion, and real-time performance.

[0003] In the prior art, the stability and collection accuracy of UAV monitoring systems in deep water areas are insufficient. Water sample collection UAVs mainly rely on single-point collection and simple sensor data processing, lacking the ability to collect samples at multiple points in deep water areas, and the data processing is mostly single-sensor data, lacking the comprehensive analysis ability of multi-sensor data fusion. Summary of the Invention

[0004] In view of the deficiencies of the prior art, the present invention makes full use of the unique advantages of multi-region collection and data fusion, aiming to achieve comprehensive multi-point water sample collection and multi-sensor data fusion processing in deep water areas, so as to improve the comprehensiveness of sampling data and the accuracy of data analysis. Specifically, a UAV monitoring system and method based on deep water multi-region water sample collection and multi-sensor data fusion are proposed.

[0005] To achieve the foregoing invention object, the present invention adopts the following solutions:

[0006] One aspect of the present invention provides a UAV monitoring system based on deep water multi-region water sample collection and multi-sensor data fusion, including: a telescopic sampling arm, a multi-sensor module, a data transmission module, and an intelligent data processing module;

[0007] A plurality of water sample collectors are integrated on the telescopic sampling arm for automatically collecting water samples at different water depths;

[0008] The multi-sensor module is used to monitor water samples in real time, accurately measure multi-sensor data of the sampling point, and transmit the multi-sensor data to the intelligent data processing module in real time; wherein, the multi-sensor module includes a multi-depth sensor, a multi-spectral camera, a microwave sensor, and a water quality sensor;

[0009] The data transmission module is used to transmit the collected multi-sensor data to the intelligent data processing module in real time by using a 4G or 5G network;

[0010] The intelligent data processing module is used to perform data fusion processing on the multi-sensor data collected, generate an analysis report, and display it through a graphical interface.

[0011] In one embodiment, the multi-depth sensor is used to monitor the depth information of water sample collection in real time;

[0012] The multi-spectral camera is used to photograph the color and reflection of the water body, and analyze the water body area and surface conditions;

[0013] The microwave sensor is used to measure the depth and flow rate of the water body;

[0014] The water quality sensor integrates multiple sensors such as pH value, turbidity, dissolved oxygen, temperature, and conductivity, and is used to monitor water quality parameters.

[0015] In one embodiment, the intelligent data processing module includes a data preprocessing module, a feature extraction module, and a data fusion module;

[0016] Among them, the data preprocessing module is used to perform normalization processing on the multi-sensor data;

[0017] The feature extraction module is used to automatically extract the features of the multi-sensor data through a multi-layer neural network, including time series features and spatial features;

[0018] The data fusion module fuses the multi-sensor data through the output layer of the neural network to generate a comprehensive water quality monitoring result. The specific steps are as follows:

[0019] Perform weighted summation on the data of different sensors. The formula is as follows:

[0020]

[0021] In the formula, Z is the comprehensive water quality monitoring result; ω i is the weight of the i-th sensor; x i is the data of the i-th sensor;

[0022] Or design multiple output nodes, and each node corresponds to a water quality parameter. The formula is as follows:

[0023]

[0024] In the formula, Y j is the output of the j-th water quality parameter, ω ij$w_{ij}$ is the weight of the $i$-th sensor for the $j$-th water quality parameter, and $f$ is the activation function; the comprehensive water quality monitoring result is obtained by weighted summation; a single output node is used to generate the comprehensive water quality monitoring result, while other output nodes are used to predict each water quality parameter respectively; the long short-term memory network (LSTM) of the time series model is used to fuse the time series data to capture the time dependence; the convolutional neural network (CNN) is used to fuse the spatial data to capture the spatial features.

[0025] In one embodiment, the intelligent data processing module further includes a data analysis module for analyzing the fused data to extract the water quality parameters, water depth, flow velocity, underwater terrain, and ecological environment of the water body. The specific steps are as follows: use the support vector machine (SVM) to classify the water quality level; estimate the depth by analyzing the signal transmission time or signal strength; correct the influence of environmental factors on the microwave signal; measure the water flow velocity using the Doppler effect; calculate the flow velocity by analyzing the frequency change of the microwave signal; construct the underwater topographic map by combining the positioning data and bathymetric data of the unmanned aerial vehicle (UAV); analyze the images taken by the multispectral camera to extract the color and reflectivity of the water body, and evaluate the water quality status and the distribution of aquatic vegetation.

[0026] In one embodiment, the intelligent data processing module further includes a report generation module that automatically generates a real-time monitoring report including water quality parameters, hydrological analysis, and ecological assessment, and displays it through a graphical user interface.

[0027] Another aspect of the present invention provides a UAV monitoring method based on deep-water multi-region water sample collection and multi-sensor data fusion, including:

[0028] Using a UAV with a retractable sampling arm to automatically collect water samples at different water depths; multiple water sample collectors are integrated on the retractable sampling arm.

[0029] Using multiple different types of sensors to monitor the water samples in real time and accurately measure the multi-sensor data at the sampling points. Among them, the multiple different types of sensors include multi-depth sensors, multispectral cameras, microwave sensors, and water quality sensors.

[0030] Transmitting the collected data to the ground control station in real time via 4G or 5G network.

[0031] The ground control station performs data fusion processing on the collected multi-sensor data, generates an analysis report, and displays it through a graphical interface.

[0032] In one embodiment, the multi-depth sensor is used to monitor the depth information of water sample collection in real time.

[0033] The multispectral camera is used to photograph the color and reflection of the water body, and analyze the water body area and surface conditions.

[0034] The microwave sensor is used to measure the depth and flow velocity of water bodies;

[0035] The water quality sensor integrates multiple sensors such as pH value, turbidity, dissolved oxygen, temperature, and conductivity, and is used to monitor water quality parameters.

[0036] In one embodiment, the ground control station performs data fusion processing on the multi-sensor data collected, generates an analysis report, and displays it through a graphical interface. It further includes:

[0037] Normalize the multi-sensor data;

[0038] Automatically extract the features of the multi-sensor data through a multi-layer neural network, including temporal features and spatial features;

[0039] Fuse the multi-sensor data through the output layer of the neural network to generate a comprehensive water quality monitoring result. The specific steps are as follows:

[0040] Perform weighted summation on the data of different sensors. The formula is as follows:

[0041]

[0042] In the formula, Z is the comprehensive water quality monitoring result; ω i is the weight of the i-th sensor; x i is the data of the i-th sensor;

[0043] Or design multiple output nodes, with each node corresponding to a water quality parameter. The formula is as follows:

[0044]

[0045] In the formula, Y j is the output of the j-th water quality parameter, ω ij is the weight of the i-th sensor for the j-th water quality parameter, f is the activation function; obtain the comprehensive water quality monitoring result through weighted summation; use one output node to generate the comprehensive water quality monitoring result, and at the same time use other output nodes to predict each water quality parameter respectively; use the long short-term memory network LSTM of the time series model to fuse the time series data and capture the time dependence; use the convolutional neural network CNN to fuse the spatial data and capture the spatial features.

[0046] In one embodiment, the ground control station performs data fusion processing on the collected multi-sensor data, generates an analysis report, and displays it through a graphical interface. Further, it includes: analyzing the fused data to extract water quality parameters, water depth, flow velocity, underwater terrain, and ecological environment of the water body. The specific steps are as follows: using a support vector machine (SVM) to classify water quality grades; estimating the depth by analyzing the signal transmission time or signal strength; correcting the influence of environmental factors on microwave signals; measuring the water flow velocity using the Doppler effect; calculating the flow velocity by analyzing the frequency change of microwave signals; constructing an underwater topographic map by combining the positioning data and bathymetric data of the unmanned aerial vehicle (UAV); analyzing the images taken by the multi-spectral camera, extracting the color and reflectivity of the water body, and evaluating the water quality status and the distribution of aquatic vegetation.

[0047] In one embodiment, the ground control station performs data fusion processing on the collected multi-sensor data, generates an analysis report, and displays it through a graphical interface. Further, it includes: automatically generating a real-time monitoring report including water quality parameters, hydrological analysis, and ecological assessment, and displaying it through a graphical user interface.

[0048] Compared with the prior art, the present invention has at least the following advantages:

[0049] (1) Multi-region collection: Integrating multiple water sample collectors on the UAV to achieve multi-point water sample collection by the UAV in deep water areas, improving the comprehensiveness and representativeness of the sampling data.

[0050] (2) Multi-sensor data fusion: Improving the accuracy and comprehensiveness of data analysis through the fusion processing of multi-sensor data.

[0051] (3) Deep water adaptability: Enhancing the stability and collection accuracy of the UAV in deep water areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0053] Figure 1 It is a schematic flow chart of the UAV monitoring system provided by a typical embodiment of the present invention; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0054] To make the objectives, technical solutions and advantages of the present invention more clear, the following provides a detailed description of the specific embodiments of the present invention with reference to the accompanying drawings. Examples of these preferred embodiments are illustrated in the accompanying drawings. The embodiments of the present invention shown in the drawings and described according to the drawings are merely exemplary, and the present invention is not limited to these embodiments.

[0055] One aspect of the present invention provides an unmanned aerial vehicle (UAV) monitoring system based on deep - water multi - area water sample collection and multi - sensor data fusion. As Figure 1 shown, it includes: a retractable sampling arm, a multi - depth sensor module, a data transmission module, and an intelligent data processing module;

[0056] A plurality of water sample collectors are integrated on the retractable sampling arm for automatically collecting water samples at different water depths;

[0057] The multi - sensor module is used to monitor water samples in real time, accurately measure multi - sensor data at the sampling point, and transmit the multi - sensor data to the intelligent data processing module in real time. Among them, the multi - sensor module includes a multi - depth sensor, a multi - spectral camera, a microwave sensor, and a water quality sensor;

[0058] The data transmission module is used to transmit the collected multi - sensor data to the intelligent data processing module in real time using a 4G or 5G network;

[0059] The intelligent data processing module is used to perform data fusion processing on the collected multi - sensor data, generate an analysis report, and display it through a graphical interface.

[0060] Among them, the retractable sampling arm can automatically collect water samples at different depths (such as 1 meter, 5 meters, 10 meters), ensuring representative water samples are obtained at different water depths. The sampling arm is made of corrosion - resistant materials, has waterproof and compressive capabilities, and can work stably in deep - water areas. The sampling process is automatically controlled by the intelligent control system of the UAV. The operator only needs to set the sampling depth and position at the ground control station, and the system will automatically complete the sampling process.

[0061] The water sample collector can be made of different materials or designs, such as using a foldable collector to reduce space occupation.

[0062] In addition, different types of sensors (such as conductivity sensors, dissolved oxygen sensors, etc.) can be used to replace the existing sensors to meet different environmental monitoring requirements.

[0063] In one embodiment, the multi - depth sensor is used to monitor the depth information of water sample collection in real time;

[0064] The multi - spectral camera is used to photograph the color and reflection of the water body, and analyze the water body area and surface conditions;

[0065] The microwave sensor is used to measure the depth and flow velocity of water bodies;

[0066] The water quality sensor integrates multiple sensors such as pH value, turbidity, dissolved oxygen, temperature, and conductivity, and is used to monitor water quality parameters.

[0067] Among them, the multi-depth sensor monitors the depth information of water sample collection in real time to ensure the accuracy of data. The accuracy of the depth sensor reaches ±0.1 m, which can accurately measure the depth information of the sampling point and transmit the data to the ground control station in real time;

[0068] The multi-spectral camera is used to photograph the color and reflection of the water body, and analyze the water area and surface conditions. It can capture spectral information in the wavelength range of 400 - 1000 nm, provide high-resolution water quality images with a resolution of 1080p. By analyzing the multi-spectral images with dedicated software, information such as the color and reflectivity of the water body can be extracted to help judge the pollution degree and area of the water body.

[0069] The microwave sensor is used to measure the depth and flow velocity of water bodies. By using microwave signals to penetrate the water surface, the measurement accuracy reaches ±0.5 m, and it can calculate the water flow velocity with an accuracy of ±0.1 m / s. Through the intelligent data processing system of the ground control station, the data of the microwave sensor is analyzed to generate real-time reports on water depth and flow velocity.

[0070] The water quality sensor integrates multiple sensors such as pH value, turbidity, dissolved oxygen, temperature, and conductivity, and monitors water quality parameters in real time. Through the intelligent data processing system of the ground control station, the data of the water quality sensor is analyzed to generate real-time reports on water quality parameters.

[0071] In one embodiment, the intelligent data processing module includes a data preprocessing module, a feature extraction module, and a data fusion module;

[0072] Among them, the data preprocessing module is used to normalize the multi-sensor data;

[0073] The feature extraction module is used to automatically extract the features of multi-sensor data through a multi-layer neural network, including temporal features and spatial features;

[0074] The data fusion module fuses the multi-sensor data through the output layer of the neural network to generate a comprehensive water quality monitoring result. The specific steps are as follows:

[0075] Perform weighted summation on the data of different sensors. The formula is as follows:

[0076]

[0077] In the formula, Z is the comprehensive water quality monitoring result; ω iis the weight of the i-th sensor; x i is the data of the i-th sensor;

[0078] Or design multiple output nodes, each corresponding to a water quality parameter. The formula is as follows:

[0079]

[0080] In the formula, Y j is the output of the j-th water quality parameter, ω ij is the weight of the i-th sensor for the j-th water quality parameter, f is the activation function; the comprehensive water quality monitoring result is obtained through weighted summation; use one output node to generate the comprehensive water quality monitoring result, and use other output nodes to predict each water quality parameter respectively; use the long short-term memory network LSTM of the time series model to fuse the time series data and capture the time dependence; use the convolutional neural network CNN to fuse the spatial data and capture the spatial features.

[0081] Among them, the neural network model: used for the fusion processing of complex data to improve the intelligence and reliability of monitoring. The neural network model can perform deep learning on multi-sensor data, automatically identify and correct the errors between different sensors, and continuously optimize the monitoring results through training data. At the same time, use the historical monitoring data to train the neural network and continuously optimize the model parameters to improve the monitoring accuracy.

[0082] In one embodiment, the intelligent data processing module further includes a data analysis module for analyzing the fused data to extract the water quality parameters, water depth, flow velocity, bottom topography, and ecological environment of the water body; the specific steps are as follows: use the support vector machine SVM to classify the water quality level; estimate the depth by analyzing the signal transmission time or signal strength; correct the influence of environmental factors on the microwave signal; measure the water flow velocity using the Doppler effect; calculate the flow velocity by analyzing the frequency change of the microwave signal; combine the positioning data and bathymetric data of the unmanned aerial vehicle to construct a bottom topographic map; analyze the images taken by the multispectral camera, extract the color and reflectivity of the water body, and evaluate the water quality status and the distribution of aquatic vegetation.

[0083] In one embodiment, the intelligent data processing module further includes a report generation module that automatically generates a real-time monitoring report including water quality parameters, hydrological analysis, and ecological assessment, and displays it through a graphical user interface.

[0084] Another aspect of the present invention provides a drone monitoring method based on deep-water multi-region water sample collection and multi-sensor data fusion, including:

[0085] Using a drone with a telescopic sampling arm to automatically collect water samples at different water depths; multiple water sample collectors are integrated on the telescopic sampling arm,

[0086] Use multiple sensors of different types to monitor water samples in real time and accurately measure multi-sensor data at the sampling point. Among them, the multiple sensors of different types include multi-depth sensors, multi-spectral cameras, microwave sensors, and water quality sensors;

[0087] Use 4G or 5G networks to transmit the collected data to the ground control station in real time;

[0088] The ground control station performs data fusion processing on the collected multi-sensor data, generates an analysis report, and displays it through a graphical interface.

[0089] Among them, the retractable sampling arm can automatically collect water samples at different depths (such as 1 meter, 5 meters, 10 meters) to ensure representative water samples are obtained at different water depths. The sampling arm is made of corrosion-resistant materials, has waterproof and compressive capabilities, and can work stably in deep water areas. The sampling process is automatically controlled by the intelligent control system of the unmanned aerial vehicle. The operator only needs to set the sampling depth and position at the ground control station, and the system will automatically complete the sampling process.

[0090] The water sample collector can be made of different materials or designs, such as using a foldable collector to reduce space occupancy.

[0091] In addition, different types of sensors (such as conductivity sensors, dissolved oxygen sensors, etc.) can be used to replace the existing sensors to meet different environmental monitoring requirements.

[0092] In one embodiment, the multi-depth sensor is used to monitor the depth information of water sample collection in real time;

[0093] The multi-spectral camera is used to photograph the color and reflection of the water body and analyze the water body area and surface conditions;

[0094] The microwave sensor is used to measure the depth and flow rate of the water body;

[0095] The water quality sensor integrates multiple sensors such as pH value, turbidity, dissolved oxygen, temperature, and conductivity, and is used to monitor water quality parameters.

[0096] Among them, the multi-depth sensor monitors the depth information of water sample collection in real time to ensure the accuracy of the data. The accuracy of the depth sensor reaches ±0.1 meter, which can accurately measure the depth information of the sampling point and transmit the data to the ground control station in real time;

[0097] The multispectral camera is used to capture the color and reflection of water bodies, and analyze the water body area and surface conditions. It can capture spectral information in the 400 - 1000nm band, provide high - resolution water quality images with a resolution of 1080p. By using dedicated software to analyze the multispectral images, information such as the color and reflectivity of the water body can be extracted to help judge the pollution degree and area of the water body.

[0098] The microwave sensor is used to measure the depth and flow velocity of water bodies. By using microwave signals to penetrate the water surface, the measurement accuracy reaches ±0.5 meters, and the water flow velocity can be deduced with an accuracy of ±0.1 m / s. Through the intelligent data processing system of the ground control station, the data of the microwave sensor is analyzed to generate real - time reports on water depth and flow velocity.

[0099] The water quality sensor integrates multiple sensors such as pH value, turbidity, dissolved oxygen, temperature, and conductivity to monitor water quality parameters in real - time. Through the intelligent data processing system of the ground control station, the data of the water quality sensor is analyzed to generate real - time reports on water quality parameters.

[0100] In one embodiment, the ground control station performs data fusion processing on the collected multi - sensor data, generates an analysis report, and displays it through a graphical interface. It further includes:

[0101] Normalize the multi - sensor data;

[0102] Automatically extract the features of multi - sensor data through a multi - layer neural network, including temporal features and spatial features;

[0103] Fuse the multi - sensor data through the output layer of the neural network to generate a comprehensive water quality monitoring result. The specific steps are as follows:

[0104] Perform weighted summation on different sensor data. The formula is as follows:

[0105]

[0106] In the formula, Z is the comprehensive water quality monitoring result; ω i is the weight of the i - th sensor; x i is the data of the i - th sensor;

[0107] Or design multiple output nodes, with each node corresponding to a water quality parameter. The formula is as follows:

[0108]

[0109] In the formula, Y j is the output of the j - th water quality parameter, ω ij$w_{ij}$ is the weight of the $i$-th sensor for the $j$-th water quality parameter, and $f$ is the activation function; the comprehensive water quality monitoring result is obtained by weighted summation; a single output node is used to generate the comprehensive water quality monitoring result, while other output nodes are used to predict each water quality parameter respectively; the long short-term memory network (LSTM) of the time series model is used to fuse the time series data to capture the time dependence; the convolutional neural network (CNN) is used to fuse the spatial data to capture the spatial features.

[0110] Among them, the neural network model: is used for the fusion processing of complex data to improve the intelligence and reliability of monitoring. The neural network model can perform deep learning on multi-sensor data, automatically identify and correct the errors between different sensors, and continuously optimize the monitoring results through training data. At the same time, the neural network is trained using historical monitoring data to continuously optimize the model parameters and improve the monitoring accuracy.

[0111] In one embodiment, the ground control station performs data fusion processing on the collected multi-sensor data, generates an analysis report, and displays it through a graphical interface, which further includes: analyzing the fused data to extract the water quality parameters, water depth, flow velocity, underwater terrain, and ecological environment of the water body; the specific steps are as follows: using the support vector machine (SVM) to classify the water quality level; estimating the depth by analyzing the signal transmission time or signal strength; correcting the influence of environmental factors on the microwave signal; measuring the water flow velocity using the Doppler effect; calculating the flow velocity by analyzing the frequency change of the microwave signal; constructing an underwater topographic map by combining the positioning data and bathymetric data of the unmanned aerial vehicle (UAV); analyzing the images taken by the multi-spectral camera to extract the color and reflectivity of the water body, and evaluating the water quality status and the distribution of aquatic vegetation.

[0112] In one embodiment, the ground control station performs data fusion processing on the collected multi-sensor data, generates an analysis report, and displays it through a graphical interface, which further includes: automatically generating a real-time monitoring report including water quality parameters, hydrological analysis, and ecological assessment, and displaying it through a graphical user interface.

[0113] The data fusion method can also adopt different algorithms, such as using machine learning algorithms to replace traditional statistical analysis methods to improve the intelligence and accuracy of data processing.

[0114] The route planning can adopt a dynamic adjustment strategy to adjust the position and quantity of the collection points according to the real-time environmental data.

[0115] A scheme of collaborative work between the ground station and the UAV can be used, where the ground station performs data processing and analysis, and the UAV is only responsible for data collection and transmission.

[0116] It should be understood that although this specification is described in terms of embodiments, not every embodiment contains only one independent technical solution. This narrative style of the specification is only for clarity. Those skilled in the art should regard the specification as a whole, and the technical solutions of each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. An unmanned aerial vehicle monitoring system based on deep-water multi-region water sampling and multi-sensor data fusion, characterized in that: include: Retractable sampling arm, multi-sensor module, data transmission module, intelligent data processing module; The retractable sampling arm is integrated with a plurality of water sample collectors for automatically collecting water samples at different water depths; The multi-sensor module is used to monitor water samples in real time, accurately measure multi-sensor data at sampling points, and transmit the multi-sensor data to the intelligent data processing module in real time; wherein the multi-sensor module includes a multi-depth sensor, a multi-spectral camera, a microwave sensor, and a water quality sensor; The data transmission module is used to transmit the collected multi-sensor data to the intelligent data processing module in real time using a 4G or 5G network; The intelligent data processing module is used to perform data fusion processing on the collected multi-sensor data, generate an analysis report, and display it through a graphical interface.

2. The drone monitoring system according to claim 1, characterized in that: The multi-depth sensor is used to monitor the depth information of water sample collection in real time; The multispectral camera is used to capture the color and reflection of the water body and analyze the area and surface conditions of the water body; The microwave sensor is used to measure the depth and flow rate of the water body; The water quality sensor integrates multiple sensors such as pH, turbidity, dissolved oxygen, temperature and conductivity to monitor water quality parameters.

3. The drone monitoring system according to claim 1, characterized in that: The intelligent data processing module includes a data preprocessing module, a feature extraction module, and a data fusion module; Among them, the data preprocessing module is used to normalize the multi-sensor data; The feature extraction module is used to automatically extract the features of multi-sensor data through a multi-layer neural network, including temporal features and spatial features; The data fusion module fuses the multi-sensor data through the output layer of the neural network to generate comprehensive water quality monitoring results. The specific steps are as follows: The weighted sum of different sensor data is as follows: Where Z is the comprehensive water quality monitoring result; ω i is the weight of the i-th sensor; x i is the data of the i-th sensor; Or design multiple output nodes, each node corresponds to a water quality parameter. The formula is as follows: Where Y j is the output of the jth water quality parameter, ω ij is the weight of the i-th sensor to the j-th water quality parameter, and f is the activation function; the comprehensive water quality monitoring results are obtained by weighted summation; one output node is used to generate the comprehensive water quality monitoring results, and other output nodes are used to predict each water quality parameter respectively; the time series model long short-term memory network LSTM is used to fuse time series data and capture time dependency; the convolutional neural network CNN is used to fuse spatial data and capture spatial features.

4. The drone monitoring system according to claim 1, characterized in that: The intelligent data processing module also includes a data analysis module, which is used to analyze the fused data and extract the water quality parameters, water depth, flow rate, underwater topography, and ecological environment of the water body; the specific steps are as follows: use support vector machine SVM to classify water quality levels; estimate depth by analyzing signal transmission time or signal strength; correct the influence of microwave signals in combination with environmental factors; measure water flow velocity using the Doppler effect; calculate flow velocity by analyzing the frequency change of microwave signals; construct an underwater topographic map by combining the positioning data and depth measurement data of the drone; analyze images taken by a multispectral camera, extract the color and reflectivity of the water body, and evaluate water quality and aquatic vegetation distribution.

5. The drone monitoring system according to claim 1, characterized in that: The intelligent data processing module also includes a report generation module that automatically generates a real-time monitoring report including water quality parameters, hydrological analysis, and ecological assessment, and displays it through a graphical user interface.

6. A drone monitoring method based on deep-water multi-region water sampling and multi-sensor data fusion, characterized in that: include: Using drones with retractable sampling arms to automatically collect water samples at different water depths; The retractable sampling arm is integrated with multiple water sample collectors. Use multiple different types of sensors to monitor water samples in real time and accurately measure multi-sensor data at sampling points. Multiple different types of sensors include multi-depth sensors, multi-spectral cameras, microwave sensors, and water quality sensors. Use 4G or 5G networks to transmit the collected data to the ground control station in real time; The ground control station performs data fusion processing on the collected multi-sensor data, generates an analysis report, and displays it through a graphical interface.

7. The drone monitoring method according to claim 6, characterized in that: The multi-depth sensor is used to monitor the depth information of water sample collection in real time; The multispectral camera is used to capture the color and reflection of the water body and analyze the area and surface conditions of the water body; The microwave sensor is used to measure the depth and flow rate of the water body; The water quality sensor integrates multiple sensors such as pH, turbidity, dissolved oxygen, temperature and conductivity to monitor water quality parameters.

8. The drone monitoring method according to claim 6, characterized in that: The ground control station performs data fusion processing on the collected multi-sensor data, generates an analysis report, and displays it through a graphical interface, further comprising: Normalize multi-sensor data; Automatically extract the features of multi-sensor data through multi-layer neural networks, including temporal features and spatial features; The multi-sensor data is fused through the output layer of the neural network to generate comprehensive water quality monitoring results. The specific steps are as follows: The weighted sum of different sensor data is as follows: Where Z is the comprehensive water quality monitoring result; ω i is the weight of the i-th sensor; x i is the data of the i-th sensor; Or design multiple output nodes, each node corresponds to a water quality parameter. The formula is as follows: Where Y j is the output of the jth water quality parameter, ω ij is the weight of the i-th sensor to the j-th water quality parameter, and f is the activation function; the comprehensive water quality monitoring results are obtained by weighted summation; one output node is used to generate the comprehensive water quality monitoring results, and other output nodes are used to predict each water quality parameter respectively; the time series model long short-term memory network LSTM is used to fuse time series data and capture time dependency; the convolutional neural network CNN is used to fuse spatial data and capture spatial features.

9. The drone monitoring method according to claim 6, characterized in that: The ground control station performs data fusion processing on the collected multi-sensor data, generates an analysis report, and displays it through a graphical interface, further including: analyzing the fused data to extract water quality parameters, water depth, flow rate, underwater topography, and ecological environment of the water body; the specific steps are as follows: using support vector machine SVM to classify water quality levels; estimating depth by analyzing signal transmission time or signal strength; correcting the influence of microwave signals in combination with environmental factors; measuring water flow velocity using the Doppler effect; calculating flow velocity by analyzing frequency changes of microwave signals; constructing an underwater topographic map in combination with the positioning data and bathymetric data of the drone; analyzing images taken by a multispectral camera, extracting the color and reflectivity of the water body, and evaluating water quality and aquatic vegetation distribution.

10. The drone monitoring method according to claim 6, characterized in that: The ground control station performs data fusion processing on the collected multi-sensor data, generates an analysis report, and displays it through a graphical interface, further including: automatically generating a real-time monitoring report including water quality parameters, hydrological analysis, and ecological assessment, and displaying it through a graphical user interface.

Images