Multi-mode small reservoir impounded water monitoring system

Abstract

The invention relates to the technical field of reservoir monitoring, and particularly discloses a multi-modal small reservoir impounded water monitoring system, which comprises a data acquisition module, a model construction module, a data monitoring module, an impounded water decision estimation module and a database module which are electrically connected with one another, the data acquisition module obtains three-dimensional topographic data and water level data of the small reservoir through unmanned aerial vehicle surveying and mapping, and the model construction module is used for constructing a three-dimensional model of the small reservoir according to the three-dimensional topographic data and the water level data; the data monitoring module is used for acquiring real-time water level data of the small reservoir, the remaining reservoir capacity of the reservoir and the water inflow speed; the water storage decision pre-estimation module calculates a pre-estimation decision of the small reservoir through a pre-trained water storage estimation model according to the water inflow speed of the small reservoir and the residual reservoir capacity of the reservoir; the objective of the invention is to solve the problem that an existing reservoir prediction model is lack of pre-estimation and judgment of future water retaining operation of a plurality of small reservoirs.

Description

Technical Field

[0001] The invention relates to the technical field of reservoir monitoring, and in particular to a multi-modal small reservoir water storage monitoring system. Background Art

[0002] Reservoir water level forecasting plays a key role in water resource management such as industry, hydropower, irrigation and water supply, as well as flood prevention and disaster reduction. Accurate water level forecasting not only helps managers formulate reasonable reservoir scheduling strategies and optimize water resource allocation, but also significantly improves flood control capabilities, thereby effectively reducing losses caused by natural disasters. As an important prerequisite for reservoir operation and scheduling, accurate water level forecasting has far-reaching significance for ensuring flood prevention and rescue, drought relief, efficient use and allocation of water resources, and maintaining and protecting the water ecological environment.

[0003] The change of reservoir water level is affected by many factors, which jointly determine the fluctuation process of reservoir water level in the future. In recent years, with the development of computer science, machine learning methods and deep learning methods for reservoir water level have developed rapidly. Usually, the historical information of rainfall, historical information of average rainfall and historical information of reservoir water level of multiple stations are taken; and based on the acquired historical information and the trained reservoir water level prediction model, the reservoir water level information at multiple moments in the future is predicted. However, the existing reservoir water level prediction is generally only for medium and large reservoirs, such as hydropower stations. There are many small reservoirs in the upstream rivers of some medium and large reservoirs; because the economic utilization value of many small reservoirs is relatively low, general management units often only perform basic water intake and water storage operations on them, and rarely carry out continuous and effective supervision of water levels. In addition, there are a large number of unmanned small reservoirs, which are at risk of being easily damaged even if water level monitoring equipment is installed. In addition, many small reservoirs are located in remote mountainous areas with complex geographical locations and difficult to reach, which further increases the difficulty of daily monitoring and management.

[0004] Therefore, when a river flows through a small reservoir, it may be impounded by the small reservoir, causing the actual water inflow to the target reservoir to be smaller than predicted, or the small reservoir may release water into the river, causing the water inflow to be larger than predicted. However, due to the lack of monitoring and prediction methods for small reservoirs, and the lack of methods to predict the water interception and storage operations of multiple small reservoirs, it is impossible to know the impact of upstream small reservoirs on the target reservoir, resulting in inaccurate water level prediction results for the target reservoir by the existing prediction model. How to predict the future water interception and storage operations of multiple small reservoirs to improve the accuracy of water level prediction for the target reservoir is an urgent problem to be solved. Summary of the invention

[0005] In view of the deficiencies in the prior art, the technical problem solved by the present invention is to provide a multi-modal small reservoir water storage monitoring system to solve the problem that the existing reservoir prediction model lacks the ability to predict and judge the future water storage operations of multiple small reservoirs.

[0006] In order to solve the above problems, the technical solution adopted by the present invention is: a multi-modal small reservoir water storage monitoring system, including a data acquisition module, a model building module, a data monitoring module, a water storage decision estimation module, and a database module electrically connected to each other; The data acquisition module acquires three-dimensional terrain data and water level data of the small reservoir through UAV mapping, and the model construction module is used to construct a three-dimensional model of the small reservoir based on the three-dimensional terrain data and water level data; The data monitoring module is used to obtain the real-time water level data of the small reservoir through one or more methods of remote sensing monitoring, drone monitoring, and water level sensor monitoring and store it in the database module, and calculate the remaining storage capacity of the reservoir based on the current water level data, and is also used to obtain the water inflow rate of the small reservoir through a radar flow meter; The water storage decision estimation module calculates the estimated decision of the small reservoir according to the water inflow rate of the small reservoir and the remaining storage capacity of the reservoir through a pre-trained water storage estimation model; The water storage estimation model is:

[0007] in and is the weight coefficient; represents the state of a small reservoir, where is the capacity state matrix of the small reservoir, is the water inflow velocity matrix; Represents the water storage and release status decision matrix of all tasks; resource allocation matrix represents the ratio of the storage space allocated to each reservoir for each inlet and outlet to the available space of all small reservoirs in the system; the set Represents N small reservoirs, set Represents the flood discharge channel of each small reservoir, the set It means a river; The total cost formula is ; The total revenue formula is ; , Respectively represent the storage and outbound time, Indicates the remaining capacity. Represents the cost required to complete the task. The cost of each drainage task is proportional to the water inflow rate and inversely proportional to the remaining capacity of the reservoir. .

[0008] The beneficial effects of this program are: There are many small reservoirs in the upstream rivers of some existing medium and large prediction target reservoirs. When the river flows through the small reservoirs, it may be impounded by the small reservoirs, resulting in a deviation between the actual water inflow and the predicted value, causing an error in the predicted value of the target reservoir, and then causing errors in the water source dispatching decision of the medium and large target reservoirs; this system obtains the water inflow of each small reservoir and the real-time water level data of the small reservoir, and then obtains the water storage decision of each small reservoir through the water storage estimation model, so that the error value can be calculated according to the water storage decision of the small reservoir, for example, X of the N small reservoirs decide to store water, and the reduction value of the predicted water inflow of the target reservoir can be obtained according to the total water storage determined by the X small reservoirs, and the reduction value can be fed back to the original prediction system to update the original estimated value to make it closer to the actual situation, thereby realizing the correction of the predicted value of the target reservoir, thereby improving the accuracy of the prediction result. In this way, not only can the uncertainty caused by the drainage and storage activities of small reservoirs upstream be effectively alleviated, but more reliable and accurate data support can be provided for the water source scheduling decisions of medium and large target reservoirs, making the scheduling and management of water resources more scientific and reasonable, effectively avoiding various problems caused by forecasting errors, and ensuring the safe and stable operation of water conservancy projects and the efficient use of water resources; it can also promote the rational allocation of water resources in the region and enhance drought and flood resistance.

[0009] Real-time water level data of small reservoirs are obtained through multimodal methods such as remote sensing monitoring, drone monitoring, and water level sensor monitoring. Remote sensing monitoring and drone monitoring are suitable for small reservoirs that are unmanned or remote and difficult to reach. For places with supervision and conditions for the arrangement of fixed facilities, high-precision water level sensors are installed to achieve continuous monitoring, thereby avoiding the loss of water level sensors and the inability to obtain real-time water level information of small reservoirs. These three monitoring methods cooperate with each other and complement each other to form a strict data acquisition network. It effectively avoids problems caused by the limitations of a single monitoring method, such as the loss or damage of water level sensors or the insufficient accuracy of remote sensing monitoring in certain special environments, and ensures the real-time, effectiveness and accuracy of water level acquisition.

[0010] Furthermore, the data monitoring module includes a remote sensing monitoring unit, an analysis and judgment unit, an unmanned aerial vehicle monitoring unit and a data processing unit. The remote sensing monitoring unit is used to communicate with the satellite to obtain remote sensing images of the small reservoir. The analysis and judgment unit is used to determine whether there is any occlusion or missing in the remote sensing image of the small reservoir. If the judgment result is yes, the first flight information is sent to the unmanned aerial vehicle monitoring unit. If the judgment result is no, the remote sensing image of the small reservoir is sent to the data processing unit; the unmanned aerial vehicle monitoring unit is used to plan a flight route according to the first flight information, and control the unmanned aerial vehicle to fly according to the flight route to obtain aerial images, and send the aerial images to the data processing unit. The data processing unit is used to calculate the current water level data of the small reservoir based on the complete remote sensing image of the small reservoir, or calculate the current water level data of the small reservoir based on the missing remote sensing image of the small reservoir and the aerial image, and calculate the remaining storage capacity of the reservoir based on the water level data of the small reservoir.

[0011] This method combines the advantages of remote sensing data and aerial photography data. Remote sensing data usually has a lower spatial resolution and a wider coverage, and can provide large-scale background information; while aerial photography data has a high spatial resolution and can provide more detailed local information. This method can combine the two data sources. When the satellite is interfered by meteorological or terrain, making it impossible to obtain all small reservoir images, the missing images can be supplemented by drone aerial photography, which can reduce the dependence on a single data source and improve data accuracy. Compared with the method of using drone aerial photography for all, it reduces the consumption of computing resources to a certain extent.

[0012] Furthermore, the drone monitoring unit is also used to set the drone aerial photography frequency according to the water inflow parameters of the small reservoir. The water inflow parameters include the water inflow speed and the water inflow speed change value. The water inflow speed is defined as 1 if it is greater than the first set threshold, and 0 if it is less than it. The water inflow speed change value is defined as 1 if it is greater than the second set threshold, and 0 if it is less than it. There are four parameter combinations: 11, 10, 01, and 00, which correspond to the first aerial photography frequency, the second aerial photography frequency, the third aerial photography frequency, and the fourth aerial photography frequency, respectively.

[0013] The greater the inflow, the greater the fluctuation of reservoir water level data; the greater the change of inflow, that is, the more unstable the inflow, the greater the fluctuation of reservoir water level data; there is a positive correlation between the inflow parameter and the fluctuation of reservoir water level; when the inflow is large but within a relatively stable range, the change of reservoir water level presents a relatively regular linear change; when the inflow is small but unstable, the impact on reservoir water level is small, and the number of drone aerial photography can be relatively reduced; therefore, the values ​​of the first aerial photography frequency, the second aerial photography frequency, the third aerial photography frequency and the fourth aerial photography frequency decrease successively, so that the drone can perform multiple aerial photography when the water level fluctuates greatly, so as to obtain more diversified data and improve the accuracy of water level data acquisition.

[0014] Furthermore, the analysis and judgment unit is also used to determine whether the proportion of the area of ​​the missing part of the remote sensing image of the small reservoir to the entire area of ​​the small reservoir is greater than a set ratio. If the judgment result is yes, the second flight information is generated based on the entire area of ​​the small reservoir. The UAV detection unit is also used to generate a planned flight route based on the second flight information, and control the UAV to fly according to the flight route to obtain a complete aerial image of the small reservoir. The data processing unit is also used to calculate the current water level data of the small reservoir based on the complete aerial image of the small reservoir.

[0015] When the remote sensing image of the entire small reservoir is seriously missing, such as exceeding 70%, the accuracy of the image data obtained through remote sensing images will decrease, which will lead to a high image error when the remote sensing image and aerial image are combined to restore the image of the small reservoir, affecting the calculated water level data; therefore, a drone is used to fully cover the target water area to obtain all images of the small reservoir, so that the image of the small reservoir is restored only by aerial images and then the water level data is obtained, ensuring the accuracy of the water level data obtained and avoiding decision-making misjudgments of the entire system model.

[0016] Furthermore, the data monitoring module also includes a water level sensor monitoring module, which is used to communicate with a water level sensor to obtain real-time water level data of a small reservoir, and the water level sensor is placed in a small reservoir with someone on duty.

[0017] Furthermore, the water level sensor adopts a radar water level monitor or a drop-in water level meter. The radar water level monitor is more suitable for water level monitoring scenarios that require non-contact, high-precision, and strong stability; the drop-in water level meter has the characteristics of easy installation, strong adaptability, and low cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a schematic diagram of the water storage and release tasks of small reservoirs in this system.

[0019] Figure 2 This is a schematic diagram of the system structure.

[0020] Figure 3 Schematic diagram of the process of finding the optimal solution of the model. DETAILED DESCRIPTION

[0021] The following is further described in detail through specific implementation methods: Example 1 This is as attached Figure 1 , Figure 2 As shown: a multi-modal small reservoir water storage monitoring system, including a data acquisition module, a model building module, a data monitoring module, a water storage decision estimation module, a database module, and a data correction module that are electrically connected to each other; The data acquisition module obtains the three-dimensional terrain data and water level data of the small reservoir through UAV mapping. The model building module is used to build a three-dimensional model of the small reservoir based on the three-dimensional terrain data and water level data; the database module is used to store the historical water level data of the small reservoir; The data monitoring module is used to obtain the real-time water level data of the small reservoir through one or more methods such as remote sensing monitoring, drone monitoring, and water level sensor monitoring, and calculate the remaining storage capacity of the reservoir based on the current water level data. It is also used to obtain the water inflow rate of the small reservoir through a radar flow meter. Specifically, the data monitoring module includes a water level sensor monitoring module, a remote sensing monitoring unit, an analysis and judgment unit, a drone monitoring unit, a water inflow monitoring unit, and a data processing unit. The water level sensor monitoring module is used to communicate with the water level sensor to obtain the water level data of the small reservoir in real time. The water level sensor is placed in a manned small reservoir. The water level sensor uses a radar water level monitor or a drop-in water level gauge. The remote sensing monitoring unit is used to communicate with the satellite to obtain the remote sensing image of the small reservoir. The analysis and judgment unit is used to judge whether the remote sensing image of the small reservoir is blocked or missing. If the judgment result is yes, the first flight information is sent to the drone monitoring unit. If the judgment result is no, the remote sensing image of the small reservoir is sent to the data processing unit; the drone monitoring unit is used to plan the flight route according to the first flight information, and control the drone to fly according to the flight route to obtain aerial images, and send the aerial images to the data processing unit. The data processing unit is used to calculate the current small reservoir water level data based on the complete small reservoir remote sensing image, or calculate the current small reservoir water level data based on the missing small reservoir remote sensing image and aerial images, and calculate the remaining storage capacity of the reservoir based on the small reservoir water level data.

[0022] The water storage decision estimation module calculates the estimated decision of the small reservoir based on the water inflow rate of the small reservoir and the remaining storage capacity of the reservoir through a pre-trained water storage estimation model.

[0023] There are many small reservoirs in the upstream rivers of some existing medium and large prediction target reservoirs. When the river flows through the small reservoirs, it may be impounded by the small reservoirs, resulting in a deviation between the actual water inflow and the predicted value, causing an error in the predicted value of the target reservoir, and then causing errors in the water source dispatching decision of the medium and large target reservoirs; this system obtains the water inflow of each small reservoir and the real-time water level data of the small reservoir, and then obtains the water storage decision of each small reservoir through the water storage estimation model, so that the error value can be calculated according to the water storage decision of the small reservoir, for example, X of the N small reservoirs decide to store water, and the reduction value of the predicted water inflow of the target reservoir can be obtained according to the total water storage determined by the X small reservoirs, and the reduction value can be fed back to the original prediction system to update the original estimated value to make it closer to the actual situation, thereby realizing the correction of the predicted value of the target reservoir, thereby improving the accuracy of the prediction result. In this way, not only can the uncertainty caused by the drainage and storage activities of small reservoirs upstream be effectively alleviated, but more reliable and accurate data support can be provided for the water source scheduling decisions of medium and large target reservoirs, making the scheduling and management of water resources more scientific and reasonable, effectively avoiding various problems caused by forecasting errors, and ensuring the safe and stable operation of water conservancy projects and the efficient use of water resources; it can also promote the rational allocation of water resources in the region and enhance drought and flood resistance.

[0024] The entire water storage estimation model in this system is composed of rivers, Small reservoirs, each of which has flood discharge channels; and use the collection Represents a small reservoir, collection Represents the flood discharge channel of each small reservoir; set Indicates a river, and uses 0 to indicate a small reservoir that stores water by itself. Indicates The first small reservoir The water discharge task of the flood discharge outlet is It is the water storage and release decision for this task. Represents the water storage and release decision configuration of all tasks in the system model. In addition, It is a binary water storage and release decision, which indicates the calculation task Either the water is stored in the reservoir or released into the river through flood discharge. For example, if , Task Will flow into the river Execute. This means that water will be stored in this reservoir, i.e. The optimization goal of this model is to maximize the profit and minimize the cost , since the cost of water storage is ,in Indicates the remaining capacity. Indicates the cost required to complete the task, parameter The value of is set according to different situations of the reservoir. The cost of draining water to the river is inversely proportional to the inflow rate and directly proportional to the residual capacity of the reservoir, which means that as the residual capacity decreases, the cost of reservoir water storage operation becomes increasingly higher and tends to infinity. Therefore, in the process of solving the optimal result of this model, it will not lead to unlimited water storage in the reservoir.

[0025] The water storage estimation model is:

[0026] in and is the weight coefficient; represents the state of a small reservoir, where is the capacity state matrix of the small reservoir, is the water inflow velocity matrix; Represents the water storage and release status decision matrix of all tasks; resource allocation matrix represents the ratio of the storage space allocated to each reservoir for each inlet and outlet to the available space of all small reservoirs in the system; the set Represents N small reservoirs, set Represents the flood discharge channel of each small reservoir, the set It means a river; The total cost formula is ; The total revenue formula is .

[0027] in , Indicates the time of entry and exit respectively.

[0028] From the water storage estimation model, we can know that the calculation of water storage and water release is is a binary variable, resource allocation weight is a continuous variable, so the process of solving the optimal solution is a mixed integer nonlinear programming problem, which is difficult to solve. In order to simplify the solution, the optimization problem is decomposed into two sub-problems: decision sub-problem and resource allocation sub-problem. , the resource allocation problem is reduced to a convex optimization problem. Figure 3 Find the process for problem solving.

[0029] To solve the computational offloading decision subproblem, this solution proposes an advanced online offloading algorithm based on multi-agent reinforcement learning, which can solve the above computational problem within milliseconds. The existing meta-heuristic search algorithm, binary search method, is used to calculate the offloading decision. It is very time-consuming to search for the optimal computing offloading decision in the action space. These algorithms require a large number of iterations to fit and cannot meet the needs of online computing decision-making in this system environment. However, this solution can quickly solve the offloading decision sub-problem.

[0030] For the resource allocation subproblem, since the resource allocation of each reservoir does not affect each other, given the unloading decision ,The space allocation problem within each small reservoir is a convex optimization problem and ,is easy to solve, for example, it can be solved using the binary search ,method.

[0031] The online unloading algorithm based on multi-agent reinforcement learning in this scheme can be modeled as a Markov map. Each small reservoir is regarded as an agent, and the calculation of water storage and drainage problem is modeled as a discrete-time Markov map. The above content can be easily expressed as an MA model, which can be defined as follows:

[0032] : The set of all intelligent agents, that is, the set of all small reservoirs; : The state set of all agents,

[0033] : The set of actions of all agents,

[0034] : The set of immediate reward functions for all agents,

[0035] : All agents obtain the probability distribution of the next state when given a state and action, In this system, the goal is to minimize the cost, From the time step The total discounted reward value to start with.

[0036]

[0037] here is the immediate reward, and the weighted system cost The relationship is as follows:

[0038] Indicates the time step at which the computational unloading ends.

[0039] Discount Factor represents the present value of future rewards, A value close to 0 will lead to "short-sighted" evaluation, that is, only the current or recent reward value is considered when making decisions, and the algorithm is very likely to fall into a local optimum; Close to 1 will lead to "foresighted" evaluation, that is, the decision-making will take into account the possible long-term reward value affected by the current action, which is conducive to maximizing the entire decision-making process or long-term benefits. In this environment, the infinite future benefits will be considered, which is beyond the scope of the current action. It is obviously unreasonable and will lead to It cannot converge, so Usually, a value close to 1 is taken, such as 0.9 or 0.95. In this embodiment, 0.9 is selected.

[0040] This implementation first solves the decision of the task through the above model , and then through the decision of the given task Calculate the optimal resource allocation weights , thus solving the entire model and obtaining the estimated decisions of each small reservoir in the water storage estimation model. The step-by-step solution simplifies the problem solving difficulty and improves the online computing efficiency of the water storage estimation model.

[0041] The same parts as Example 1 are not repeated here. The difference is that the UAV monitoring unit is also used to set the UAV aerial photography frequency according to the water inflow parameters of the small reservoir. The water inflow parameters include the water inflow speed and the water inflow speed change value. The water inflow speed is defined as 1 if it is greater than the first set threshold value, and 0 if it is less than it. The water inflow speed change value is defined as 1 if it is greater than the second set threshold value, and 0 if it is less than it. There are four parameter combinations: 11, 10, 01, and 00, which correspond to the first aerial photography frequency, the second aerial photography frequency, the third aerial photography frequency, and the fourth aerial photography frequency, respectively.

[0042] The greater the inflow, the greater the fluctuation of reservoir water level data; the greater the change of inflow, that is, the more unstable the inflow, the greater the fluctuation of reservoir water level data; there is a positive correlation between the inflow parameter and the fluctuation of reservoir water level; when the inflow is large but within a relatively stable range, the change of reservoir water level presents a relatively regular linear change; when the inflow is small but unstable, the impact on reservoir water level is small, and the number of drone aerial photography can be relatively reduced; therefore, the values ​​of the first aerial photography frequency, the second aerial photography frequency, the third aerial photography frequency and the fourth aerial photography frequency decrease successively, so that the drone can perform multiple aerial photography when the water level fluctuates greatly, so as to obtain more diversified data and improve the accuracy of water level data acquisition.

[0043] The analysis and judgment unit is also used to judge whether the proportion of the area of ​​the missing part of the remote sensing image of the small reservoir to the entire area of ​​the small reservoir is greater than a set ratio. If the judgment result is yes, the second flight information is generated according to the entire area of ​​the small reservoir. The UAV detection unit is also used to generate a planned flight route according to the second flight information, and control the UAV to fly according to the flight route to obtain a complete aerial image of the small reservoir. The data processing unit is also used to calculate the current water level data of the small reservoir based on the complete aerial image of the small reservoir.

[0044] When the remote sensing image of the entire small reservoir is seriously missing, such as exceeding 70%, the accuracy of the image data obtained through remote sensing images will decrease, which will lead to a high image error when the remote sensing image and aerial image are combined to restore the image of the small reservoir, affecting the calculated water level data; therefore, a drone is used to fully cover the target water area to obtain all images of the small reservoir, so that the image of the small reservoir is restored only by aerial images and then the water level data is obtained, ensuring the accuracy of the water level data obtained and avoiding decision-making misjudgments of the entire system model.

[0045] The above is only an embodiment of the present invention, and the common knowledge such as the known specific structure and characteristics in the scheme is not described in detail here. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A multi-modal small reservoir water storage monitoring system, characterized by: It includes a data acquisition module, a model building module, a data monitoring module, a water storage decision estimation module, and a database module which are electrically connected to each other; The data acquisition module acquires the three-dimensional terrain data and water level data of the small reservoir through UAV mapping; the model construction module is used to construct a three-dimensional model of the small reservoir based on the three-dimensional terrain data and water level data; the database module is used to store the historical water level data of the small reservoir; The data monitoring module is used to obtain the real-time water level data of the small reservoir through one or more methods of remote sensing monitoring, drone monitoring, and water level sensor monitoring, and calculate the remaining storage capacity of the reservoir based on the current water level data, and is also used to obtain the water inflow rate of the small reservoir through a radar flow meter; The water storage decision estimation module calculates the estimated decision of the small reservoir according to the water inflow rate of the small reservoir and the remaining storage capacity of the reservoir through a pre-trained water storage estimation model; The water storage estimation model is: in and is the weight coefficient; represents the state of a small reservoir, where is the capacity state matrix of the small reservoir, is the water inflow velocity matrix; Represents the water storage and release status decision matrix of all tasks; resource allocation matrix represents the ratio of the storage space allocation of small reservoirs for each inlet and outlet to the available space of all small reservoirs in the system; the set Represents N small reservoirs, set Represents the flood discharge channel of each small reservoir, the set It means a river; The total cost formula is ; The total revenue formula is ; , Respectively represent the storage and outbound time, Indicates the remaining capacity. Represents the cost required to complete the task. The cost of each drainage task is proportional to the water inflow rate and inversely proportional to the remaining capacity of the reservoir. .

2. A multi-modal small reservoir water storage monitoring system according to claim 1, characterized in that: The data monitoring module includes a remote sensing monitoring unit, an analysis and judgment unit, an unmanned aerial vehicle monitoring unit and a data processing unit. The remote sensing monitoring unit is used to communicate with a satellite to obtain remote sensing images of a small reservoir. The analysis and judgment unit is used to determine whether there is any occlusion or missing in the remote sensing image of the small reservoir. If the judgment result is yes, the first flight information is sent to the unmanned aerial vehicle monitoring unit. If the judgment result is no, the remote sensing image of the small reservoir is sent to the data processing unit; the unmanned aerial vehicle monitoring unit is used to plan a flight route according to the first flight information, and control the unmanned aerial vehicle to fly according to the flight route to obtain aerial images, and send the aerial images to the data processing unit. The data processing unit is used to calculate the current water level data of the small reservoir based on the complete remote sensing image of the small reservoir, or calculate the current water level data of the small reservoir based on the missing remote sensing image of the small reservoir and the aerial image, and calculate the remaining storage capacity of the reservoir based on the water level data of the small reservoir.

3. A multi-modal small reservoir water storage monitoring system according to claim 2, characterized in that: The drone monitoring unit is also used to set the drone aerial photography frequency according to the water inflow parameters of the small reservoir. The water inflow parameters include the water inflow speed and the water inflow speed change value. The water inflow speed is defined as 1 if it is greater than the first set threshold, and 0 if it is less than the first set threshold. The water inflow speed change value is defined as 1 if it is greater than the second set threshold, and 0 if it is less than the second set threshold. There are four parameter combinations: 11, 10, 01, and 00, which correspond to the first aerial photography frequency, the second aerial photography frequency, the third aerial photography frequency, and the fourth aerial photography frequency, respectively.

4. A multi-modal small reservoir water storage monitoring system according to claim 2, characterized in that: The analysis and judgment unit is also used to judge whether the proportion of the area of ​​the missing part of the remote sensing image of the small reservoir to the entire area of ​​the small reservoir is greater than a set ratio. If the judgment result is yes, the second flight information is generated according to the entire area of ​​the small reservoir. The UAV detection unit is also used to generate a planned flight route according to the second flight information, and control the UAV to fly according to the flight route to obtain a complete aerial image of the small reservoir. The data processing unit is also used to calculate the current water level data of the small reservoir based on the complete aerial image of the small reservoir.

5. A multi-modal small reservoir water storage monitoring system according to claim 2, characterized in that: The data monitoring module also includes a water level sensor monitoring module, which is used to communicate with a water level sensor to obtain small reservoir water level data in real time. The water level sensor is placed in a small reservoir with someone on duty.

6. A multi-modal small reservoir water storage monitoring system according to claim 5, characterized in that: The water level sensor adopts a radar water level monitor or an immersion water level meter.

Images