Water conservancy monitoring system and method based on dual-spectrum sensor carried by unmanned aerial vehicle

Abstract

The invention discloses a water conservancy monitoring method based on a dual-spectrum sensor carried by an unmanned aerial vehicle, and relates to the technical field of water resource monitoring and management. According to the invention, by combining the multispectral sensor and the hyperspectral sensor, the flexibility and high-resolution data acquisition advantages of the unmanned aerial vehicle and the unique capability of multispectral and hyperspectral remote sensing in information acquisition and analysis are fully exerted, comprehensive and fine monitoring of the hydraulic engineering is realized, and the monitoring efficiency and precision are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of water resource monitoring and management, and particularly relates to a water conservancy project monitoring system and method for determining a multi-spectral and hyperspectral sensor carried by a UAV. BACKGROUND

[0002] Water conservancy projects play a crucial role in the rational allocation of water resources, flood control, drought resistance, and power generation. Timely and accurate monitoring of the operation status of water conservancy projects is of great significance for ensuring their safe and stable operation and preventing disasters. Traditional water conservancy project monitoring methods, such as manual inspection, have the disadvantages of low efficiency, strong subjectivity, difficulty in covering large areas, and difficulty in detecting hidden problems.

[0003] With the development of remote sensing technology, multi-spectral and hyperspectral remote sensing has gradually increased in the application of water conservancy project monitoring. Multi-spectral remote sensing can obtain information of ground objects in multiple specific bands, and through the analysis of different band combinations, it can identify different ground object types such as water bodies, vegetation, and soil, and has certain advantages in water conservancy project water area monitoring and vegetation coverage analysis. Hyperspectral remote sensing can obtain continuous and fine spectral information, which can more accurately identify and analyze the composition and characteristics of substances. For example, in water quality monitoring, it can detect various pollutants and nutrients in water. However, existing multi-spectral and hyperspectral remote sensing monitoring technologies still have some problems in water conservancy applications. Satellite remote sensing has a wide coverage, but the spatial resolution is limited, making it difficult to capture detailed information in water conservancy projects, and the revisit period is long, which cannot provide timely dynamic change information of the project. Ground monitoring stations can obtain accurate local information, but the monitoring range is limited, making it difficult to achieve comprehensive monitoring of large-area water conservancy projects.

[0004] The rise of UAV technology provides a new way to solve the above problems. UAVs have the advantages of flexible operation, low-altitude flight, and high-resolution image acquisition, and can quickly reach the designated area for data collection according to the monitoring requirements. However, the current combination of UAVs with multi-spectral and hyperspectral sensors for water conservancy project monitoring is not perfect, and there are problems such as low data collection efficiency, inaccurate data processing and analysis, and incomplete monitoring functions. Therefore, it is of great practical significance to develop an efficient, accurate, and comprehensive water conservancy project monitoring system and method based on UAVs carrying multi-spectral and hyperspectral sensors. SUMMARY

[0005] The present application aims to provide a water conservancy project monitoring system and method based on UAVs carrying multi-spectral and hyperspectral sensors to solve the problems of low monitoring efficiency, insufficient accuracy, and incomplete functions in existing water conservancy project monitoring technologies, and to achieve efficient, accurate, and comprehensive monitoring of water conservancy projects.

[0006] By combining hyperspectral and multispectral sensors, the following problems are solved: ① Solve the problem of lack of downstream population distribution data in small reservoir life loss risk assessment, and propose a dam break risk population estimation model based on Sentinel-2A remote sensing image. Taking a reservoir as an example, the supervised classification method is used to quickly extract the rural residential building information in the remote sensing image, and combined with the dam flood evolution result, whether the flood submerges the building is taken as the criterion, and the maximum likelihood method, support vector machine method, Mahalanobis distance method and minimum distance method are compared; ② For the problems of low accuracy and precision in the application of remote sensing image classification and synthetic aperture radar interferometric deformation measurement technology, and the problems of not being deeply integrated with engineering risk assessment; In view of the application requirements of water conservancy project construction and operation management, the business application path of water conservancy project satellite remote sensing monitoring technology is discussed in three typical application scenarios of construction progress supervision of in-construction project, accurate monitoring of in-operation project deformation and risk source identification and early warning, and the technical bottlenecks of insufficient quantitative remote sensing monitoring and insufficient multi-scale collaborative monitoring of "space-ground" are analyzed, and the development direction of water conservancy project remote sensing monitoring technology application is prospected. This paper mainly studies the risk source identification, and the remote sensing monitoring of water conservancy project risk source identification and early warning can be divided into three steps of "detection-monitoring-early warning"; ③ In order to solve the interactive influence relationship between human activities and engineering flood control safety, a human activity risk evaluation method and model of in-operation water conservancy project is proposed, and a reservoir is taken as an example. The evaluation model: considering the influence of human activity function division, human activity type and human activity area on river flow capacity, a human activity interference index model is established to analyze and evaluate the human activities in water conservancy project and downstream flood passage channel. Weight determination: based on the principle of analytic hierarchy process, the influence weight of each function division and each type of human activity is established.

[0007] The beneficial effects of the present application are:

[0008] By combining multispectral and hyperspectral sensors, the flexibility and high-resolution data acquisition advantages of unmanned aerial vehicles are fully utilized, and the unique capabilities of multispectral and hyperspectral remote sensing in information acquisition and analysis are fully utilized, so that the water conservancy project is comprehensively and finely monitored, and the monitoring efficiency and accuracy are improved. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creating laborious work.

[0010] Figure 1The preparation flowchart of the water conservancy monitoring system and method for determining the unmanned aerial vehicle carrying multi-spectral and hyperspectral sensors provided by the embodiment of the present application is shown in the figure;

[0011] Figure 2 The data processing flowchart of the embodiment of the present application is shown in the figure;

[0012] Figure 3 The main flowchart of remote sensing image supervised classification is shown in the figure;

[0013] Figure 4 The framework diagram of the monitoring automation system is shown in the figure; Figure 5 The technical roadmap of the multi-spectral and hyperspectral water conservancy monitoring system is shown in the figure. DETAILED DESCRIPTION

[0014] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0015] Embodiment 1: Reservoir monitoring

[0016] The flight task planning, data acquisition, data preprocessing, water area monitoring, reservoir monitoring and other engineering applications described in this embodiment have great significance in engineering.

[0017] In this embodiment, flight task planning: for a large reservoir, the ground control station uses professional software to plan the flight route of the unmanned aerial vehicle according to the shape, area and surrounding terrain of the reservoir. The flight height is set to 100 meters, the flight speed is 5 meters per second, the data acquisition time interval is 1 second, and the image overlap rate is 70%. The flight route covers the entire reservoir water area, the dam and the vegetation area within a certain range around the reservoir.

[0018] Data acquisition: the unmanned aerial vehicle carrying multi-spectral and hyperspectral sensors flies according to the preset route, the multi-spectral sensor acquires image data of four wavebands of blue light, green light, red light and near-infrared light, and the hyperspectral sensor acquires continuous spectral information with a spectral resolution of 1 nanometer and a spectral range of 400-1000 nanometers. During the flight, the unmanned aerial vehicle stabilizes the flight through the GPS and inertial navigation system to ensure accurate data acquisition by the sensors.

[0019] Data preprocessing: after the ground control station receives the data, the multi-spectral data is geometrically corrected and radiometrically corrected, and the features of water bodies and ground objects are enhanced through band combination; the hyperspectral data is processed for noise removal, spectral smoothing and baseline correction, etc.

[0020] Monitoring and Analysis of Water Conservancy Project Elements

[0021] Water area monitoring: Using a threshold-based image segmentation algorithm, multispectral images were processed to accurately extract the water area boundaries of the reservoir. Comparison with historical data revealed a slight increase in the reservoir's recent water area, which further analysis suggests may be due to increased inflow caused by recent heavy rainfall.

[0022] Water quality monitoring: By establishing a quantitative relationship model between water quality parameters and hyperspectral reflectance, the chlorophyll content and turbidity in the reservoir water were retrieved. The results showed that the chlorophyll content in some areas was slightly increased, which may indicate a trend towards eutrophication and requires further attention.

[0023] Embankment condition monitoring: The embankment surface was analyzed using multispectral and hyperspectral data. Image recognition technology did not reveal any obvious cracks or collapses, but hyperspectral data showed that the soil moisture in some parts of the embankment was high, indicating a potential risk of seepage, requiring further inspection.

[0024] Vegetation health monitoring: The normalized vegetation index (NDVI) of the vegetation around the reservoir was calculated. It was found that the NDVI value of vegetation in some areas was low, indicating that the vegetation growth was not good and may be affected by pests, diseases or drought. Corresponding protection measures need to be taken.

[0025] Data storage and management: The processed and analyzed data is stored in the database, and classified and labeled according to monitoring time, monitoring area, data type, etc., to facilitate subsequent query and comparative analysis.

[0026] Early warning and decision support: Based on the monitoring results, the system did not issue any early warning information, but it provided the reservoir management department with suggestions on increasing the frequency of water quality monitoring, rationally allocating water resources, and preventing and controlling diseases and pests in vegetation in response to the water quality and vegetation problems found.

[0027] Example 2: River Monitoring

[0028] The river monitoring described in this embodiment has significant implications for engineering applications such as national economic development, water supply for upstream and downstream residents, dam construction, and flood control monitoring.

[0029] In this embodiment, flight mission planning is as follows: For a relatively long river, it is divided into multiple monitoring segments, and different flight routes are planned for each monitoring segment at the ground control station. Based on the width and depth of the river, the flight altitude is set at 80 meters, the flight speed at 6 meters per second, the data acquisition interval at 0.8 seconds, and the image overlap rate at 65%. This ensures that the UAV can collect data comprehensively along the river's course.

[0030] Data Acquisition: During the drone's flight, multispectral and hyperspectral sensors work simultaneously to acquire data on the river and surrounding area. The multispectral sensor clearly records the river's boundaries and surrounding terrain features, while the hyperspectral sensor performs detailed detection of the river's water quality.

[0031] Data preprocessing: Perform routine preprocessing operations on the collected data to remove noise, correct errors, and improve data quality.

[0032] Monitoring and Analysis of Water Conservancy Project Elements

[0033] Water area monitoring: Using multispectral imagery, the water area of ​​the river was accurately determined through edge detection algorithms. Comparison of data from different monitoring sections revealed that some river sections experienced channel narrowing, possibly due to siltation or human activities.

[0034] Water quality monitoring: By inverting the concentration of pollutants in the river using hyperspectral data, it was found that the concentration of pollutants in a section of the river near an industrial cluster exceeded the standard, and relevant departments should be notified in a timely manner for investigation and treatment.

[0035] Embankment condition monitoring: Analysis of multispectral and hyperspectral data revealed vegetation damage and soil erosion on some riverbank embankments, affecting their stability and requiring prompt repair and reinforcement.

[0036] Vegetation health monitoring: Monitoring of vegetation around the river revealed that vegetation growth was good in some areas, but vegetation coverage was reduced in other areas due to human activities, requiring strengthened ecological protection measures.

[0037] Data storage and management: Data from each monitoring segment is categorized and stored in a database, with detailed indexes to facilitate subsequent comparison and analysis of data from different river segments.

[0038] Early warning and decision support: The system issues early warning information for water quality exceeding standards and potential safety hazards to dikes, while providing river management departments with decision-making suggestions for formulating dredging plans, strengthening industrial pollution supervision, carrying out dike repair projects, and strengthening riverbank ecological protection.

[0039] The above embodiments are only some implementations of the present invention. In practical applications, they can be flexibly adjusted and optimized according to the characteristics and monitoring needs of different water conservancy projects. Through the monitoring system and method of the present invention, water conservancy projects can be effectively monitored in a comprehensive and high-precision manner, providing a reliable guarantee for the safe operation and scientific management of water conservancy projects.

[0040] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

[0041] Based on the overall framework of the automated monitoring system, the automated monitoring data acquisition system adopts a hierarchical distributed intelligent network structure, divided into two levels of monitoring. The first level of monitoring involves connecting various monitoring sensors distributed across different buildings to their respective measurement control units (MCUs) for primary monitoring. The second level of monitoring connects the MCUs distributed across different locations to the monitoring and management center station within the power plant's integrated management building. The data acquisition layer utilizes the measurement control units to measure the sensors or manually input data, completing the control and management of the data acquisition layer and data aggregation, thereby achieving data acquisition, management, and analysis.

[0042] Comparison methods: The comparison methods between manual and automated testing mainly include qualitative and quantitative analysis. Commonly used qualitative analysis methods include process line analysis, and quantitative analysis methods include analysis of variance.

[0043] Process line analysis method: This method selects automated and manual measurements at the same time and frequency at a specific measuring point, plots the process lines for both automated and manual measurements, and compares the regularity and variation of the measurements to preliminarily assess the performance of the automated system, particularly its long-term stability. Compared to manual observation, automated systems monitor more frequently, resulting in more "spurts" in the automated measurement process lines, while manual measurement process lines are smoother. This discrepancy can easily affect the judgment of the accuracy and stability of the automated system. To more accurately analyze the results of manual comparisons, automated measurements taken at the same time as the manual measurements can be extracted for process line analysis.

[0044] Analysis of variance (ANOVA): ANOVA assumes that the behavior parameters of hydraulic structures remain constant over a short period. Based on this assumption, both automated monitoring and manual observation involve repeated measurements of a known effect, typically 3-6 times. Repeated measurements help reduce random errors in manual observation and make the expected value of the manual measurement closer to the true value. Selecting the same time and number of measurements from automated monitoring and manual observation at a certain monitoring location, we form automated measurement sequences and manual measurement sequences containing n data samples, respectively. Let the automated measurement value at a certain moment be Xzi, and the manual measurement value be Xri, then the difference δi between the two is...

[0045]

[0046] The standard deviation σz of automated measurements and the standard deviation σr of manual measurements are:

[0047]

[0048]

[0049]

[0050] In the formula: σ is the root mean square error between automated and manual measurements, Xz is the average value of the automated measurement sequence, and Xr is the average value of the manual measurement sequence.

[0051] The standard for controlling the difference between automated and manual measurements is as follows:

[0052]

[0053] If the difference between automated and manual measurements meets the above control standards, it indicates that the automated and manual measurements are basically stable. If the measured values ​​exceed the above standards, it should be checked whether the data is missing or incorrect, whether the instruments and equipment have been calibrated, and whether the measurement steps are correct and meet the requirements. If necessary, a retest should be conducted.

Claims

1. A water conservancy monitoring method based on a dual-spectral sensor mounted on an unmanned aerial vehicle (UAV), characterized in that: Data collection and monitoring of target areas are conducted via satellites or drones equipped with multispectral and hyperspectral remote sensing sensors and loudspeakers using the DJL Matrice 4E. Satellite remote sensing technology acquires large-scale, high-resolution data, which, combined with hyperspectral imagers such as Gaofen-5, enables comprehensive observation of water environment changes and water pollution, generating water quality parameter inversion results and monitoring analysis reports. Drones equipped with multispectral and hyperspectral sensors complete panoramic pollution inspections of water sources, identifying indicators such as phytoplankton and algae, and quickly responding to pollution source tracing needs. Smart water quality automatic monitoring stations integrate water quality, hydrological, and meteorological indicators, achieving automatic anomaly identification, remote diagnosis, and operation and maintenance through AI-powered stations. Meanwhile, the groundwater online monitoring system adopts low-flow, undisturbed sampling technology, covering 32 conventional and 18 non-conventional indicators, complying with multiple industry standards. The full-spectrum water quality monitoring system utilizes 150nm-900nm band modeling, providing second-level response to monitor multiple parameters such as COD and ammonia nitrogen, without secondary pollution. Underwater monitoring achieves precise observation of aquatic ecosystems through fish and phytoplankton monitoring.

2. The water conservancy monitoring method based on a dual-spectral sensor mounted on an unmanned aerial vehicle (UAV) according to claim 1, characterized in that: Leveraging cloud platforms such as Google Earth Engine (GEE), efficient storage, rapid processing, and automated analysis of massive amounts of multi-source remote sensing data were achieved. Addressing the limitation of traditional NDWI (Normalized Difference Water Index) in monitoring flood inundation frequency in floodplains by neglecting topographic influence, a novel topographic-multispectral hybrid model was constructed by combining high-precision LiDAR DEM (Digital Elevation Model) with Sentinel-2 multi-temporal imagery, including: When classifying unknown pixels using a model, the sample type is first determined based on the research objective and image range. Training samples are then selected to train the model. The trained model is then used to classify other images, assigning pixels to the sample type with the highest similarity according to the classification method. The classification is then completed and the results are output. A confusion matrix is ​​used to verify the extraction accuracy. This involves visually interpreting remote sensing images to obtain true regions of interest (ROIs) on the ground as validation samples. The positions and classification results of the training samples and validation samples are compared to evaluate the performance of the classification model.

3. The water conservancy monitoring method based on a dual-spectral sensor mounted on an unmanned aerial vehicle (UAV) according to claim 2, the final confusion matrix report includes the following evaluation index: A = NT / N Where: A is the total classification accuracy, NT is the number of correctly classified pixels, and N is the total number of all true reference pixels; In the formula: Ka is the Kappa coefficient, Xk is the sum of the diagonals of the confusion matrix, Nk is the total number of true reference pixels in class k, Tk is the total number of pixels correctly classified in class k, and K is the number of classification types; Ec=Ne / Ni Eo=Nm / Nr Pp=Nt / Nr Pu=Nj / Ni In the formula: Ec is the misclassification error, Ne is the number of pixels of a certain classification type that are incorrectly classified into other types, Ni is the total number of pixels of a certain classification type, Eo is the omission error, Ne is the number of pixels of a certain classification type that are incorrectly classified into other types, Nr is the total number of pixels of the class of interest of a certain classification type, Pp is the mapping accuracy, Nt is the number of pixels correctly classified of a certain classification type, Pu is the user accuracy, and Nj is the number of pixels correctly classified of a certain classification type.

4. The water conservancy monitoring method based on a dual-spectral sensor mounted on an unmanned aerial vehicle as described in claim 1, characterized in that, By identifying the characteristic absorption bands of specific pollutants using hyperspectral data, combining spatial context information from multispectral data to map the spatial distribution of pollutants, and fusion of high spatiotemporal resolution data to track the diffusion path of pollution plumes, and combining hydrodynamic models and wind direction analysis to help locate potential pollution emission sources, the study combined multispectral imagery and radar altimetry data, and relied on relevant platforms and models to extract water body information of the South Four Lakes from 2000 to 2020 and lake levels from 2003 to 2009, and constructed a water volume estimation model. It also analyzed the dynamic changes in water area, water level, and water volume in the north and south lake areas of the lake, as well as the natural and human activity driving factors of water volume changes. At the same time, the concentration values ​​of each parameter retrieved were spatially interpolated and visualized by GIS to generate a single parameter distribution map. Furthermore, a comprehensive water quality index was constructed to comprehensively evaluate the overall water quality and map different regions to show the water quality levels of different areas.

5. The water conservancy monitoring method based on a dual-spectral sensor mounted on an unmanned aerial vehicle (UAV) according to claim 4, characterized in that: HEC-RAS is a hydraulic calculation software developed by the Hydrographic Engineering Center of the U.S. Army Corps of Engineers. It is powerful, user-friendly, provides high-accuracy simulation results, is easy to use, and can interact with GIS software, making it widely used in flood simulation and analysis. HEC-RAS includes one-dimensional and two-dimensional models, with the two-dimensional model offering higher accuracy; therefore, this paper selects the two-dimensional model to simulate dam-break floods. During dam-break floods, the flow pattern is unsteady. HEC-RAS, for two-dimensional unsteady flow hydrodynamic simulation, uses the simplified Navier-Stokes equations (shallow water equations) to describe the fluid motion.

6. The water conservancy monitoring method based on a dual-spectral sensor mounted on an unmanned aerial vehicle (UAV) according to claim 5, characterized in that: The simplified Navier-Stokes equations (shallow water equations) are used to describe the fluid motion. The continuity equation and momentum equation are as follows: Where: H is the water surface elevation, t is time, V is the flow velocity, h is the water depth, q is the lateral inflow, g is the gravitational acceleration, v1 is the horizontal kinematic viscosity, Cf is the riverbed bottom roughness, f is the Coriolis coefficient, and k is the vertical unit vector. Based on the building area extracted from remote sensing imagery and the number of permanent residents at the fifth-level administrative division (village level), the population density of the study area was calculated. Then, a population grid was established, and risk population statistics were performed for each grid. The calculation primarily used whether floodwaters reached the houses as the criterion; floodwaters reaching the houses were considered to create a risk population, while those not were not. The final population estimation model is as follows: In the formula: PAR represents the total number of people at risk; n represents the total number of villages involved in the risk area; Pi represents the total number of people at risk in the i-th village within the study area; denoted as the number of flood-inundated grids in the i-th village; a0 indicates whether the flood has inundated buildings, a0=1 means the buildings have been inundated, a0=0 means the buildings have not been inundated; Sj is the area of ​​the buildings in the j-th flood-inundated grid in the i-th village; Di is the population density of the i-th village; SBUi is the total area of ​​buildings in the i-th village; PVIi is the total resident population of the i-th village.

7. The water conservancy monitoring method based on a dual-spectral sensor mounted on an unmanned aerial vehicle as described in claim 4, characterized in that, Includes the following steps: The identification and early warning of water conservancy project risk sources based on remote sensing monitoring can be divided into three steps: "detection-monitoring-early warning". First, a wide-area detection of large-scale deformation and important geological and geomorphological landmarks is carried out through multi-source satellite collaboration using high-resolution optical and SAR data. The spectral characteristics of optical remote sensing and the polarization and interferometric characteristics of microwave remote sensing are fully utilized. Based on the land cover classification features obtained by multiple methods such as neural networks, potential landslide bodies, large deformation zones and other target risk areas within the project management scope are identified and extracted. Secondly, remote sensing monitoring information such as backscattering characteristics and texture characteristics of the target risk area is analyzed. Multi-source remote sensing interpretation, including change detection by optical remote sensing and InSAR time series analysis by radar remote sensing, is used to conduct fine monitoring of changes in engineering structures. Combined with UAV aerial survey and ground monitoring verification, the changes in land cover and deformation of the target risk area are accurately tracked and monitored. Finally, based on engineering experience and early warning models, monitoring thresholds are set to conduct early warning analysis and expert evaluation of real-time monitoring results, providing support for decision-making.

8. The water conservancy monitoring method based on a dual-spectral sensor mounted on an unmanned aerial vehicle (UAV) according to claim 7, characterized in that: This includes comprehensively considering the impact of functional zones where human activities are located, types of human activities, and the area of ​​human activities on the river's flow capacity, establishing a human activity disturbance index model, and analyzing and evaluating human activities within water conservancy projects and downstream flood channels.

9. The water conservancy monitoring method based on a dual-spectral sensor mounted on an unmanned aerial vehicle (UAV) according to claim 8, characterized in that: The human activity disturbance index is calculated using the following formula: In the formula: I is the human activity interference index; aibixi is the interference intensity of the i-th type of human activity, where ai and bi are the influence weights, ai is determined according to the area where different human activity patches are located, bi is determined according to the degree of interference of human activity type to flood control of water conservancy project, xi is the area of ​​the i-th type of human activity patch; x is the total area of ​​the monitoring area. Weight determination: Based on the principles of the analytic hierarchy process (AHP), the impact weights of each functional zone and type of human activity are established. According to the criteria for defining the management and protection areas, the degree of interference of human activities on flood control of water conservancy projects within the management area is greater than that within the protection area. Therefore, the principle for assigning AI values ​​is that the scope of management > the scope of protection. After literature review and expert consultation, the weights of human activity impact on the scope of management and the scope of protection are determined to be 0.6 and 0.3, respectively. The impact weight *bi* of different types of human activities is determined based on their degree of impact on flood control in water conservancy projects. The impact of human activities on flood control in water conservancy projects is extremely complex and requires comprehensive consideration of multiple factors. Through literature review and expert consultation, the impacts of different types of human activities are compared and analyzed. Points are assigned according to the degree of impact, with scores ranging from 0 to 100. The individual score for each type of human activity is divided by the total score of all human activities to obtain the impact weight of each human activity on flood control in water conservancy projects.

Images