Land resource environment monitoring control system

Abstract

The invention relates to the technical field of intelligent control, and discloses a land resource environment monitoring control system comprising a monitoring area spatial data acquisition module used for acquiring spatial data of a monitoring area; the unmanned aerial vehicle data acquisition module is used for acquiring operation parameters and operation states of the unmanned aerial vehicle; the monitoring task analysis module is used for analyzing the monitoring task to obtain a plurality of operable sub-tasks; the data structuring module is used for constructing structured data comprising a structural dimension, and basic structural elements contained in the structural dimension are objects; the mode recognition module is used for performing feature recognition on the structured data through a flight control model to obtain flight control strategies of all the unmanned aerial vehicles for executing the monitoring tasks; the method is suitable for complex monitoring tasks of land resource environment changes caused by city expansion and infrastructure construction to automatically generate the unmanned aerial vehicle flight plan with the optimal monitoring effect.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and more specifically, to a land resource environment monitoring and control system. Background Art

[0002] UAVs have broad application prospects in land resource and environmental monitoring tasks. Remote sensing can replace a large amount of manual mapping and zoning identification work. In order to efficiently complete the monitoring task, it is necessary to plan the optimal flight route for the UAV. Traditional route planning methods mainly rely on manual design or heuristic algorithms, which are difficult to cope with increasingly complex and changing task requirements, especially for changes in land resource environment caused by urban expansion and infrastructure construction. First, the remote sensing environment is affected by construction, and the impact of these human activities on land resource environment is high-frequency dynamic changes, resulting in high-frequency dynamic changes in task requirements and environment, which further increases the difficulty of route planning. Summary of the invention

[0003] The present invention provides a land resource and environment monitoring and control system to solve the technical problem that the traditional route planning method in the related technology mainly relies on manual design or heuristic algorithm, which is difficult to cope with the monitoring needs of land resource and environment changes caused by urban expansion and infrastructure construction.

[0004] The present invention provides a land resource environment monitoring and control system, comprising:

[0005] The monitoring area spatial data acquisition module is used to collect spatial data of the monitoring area.

[0006] UAV data collection module, which is used to collect the operating parameters and operating status of the UAV;

[0007] A monitoring task parsing module, which is used to parse the monitoring task to obtain multiple operable subtasks;

[0008] The data structuring module is used to construct structured data including a structural dimension. The basic structural elements contained in this structural dimension are objects. The structuring methods include:

[0009] Extract objects that can be associated with specific features;

[0010] Establishing a mapping between the extracted objects and their associated specific features;

[0011] For any two objects, if there is a correlation between the associated specific features related to the monitoring task or there is a correlation between the objects related to the monitoring task, an object relationship is established between the two objects;

[0012] The pattern recognition module is used to perform feature recognition on structured data through the flight control model to obtain the flight control strategy used by all drones to perform monitoring tasks.

[0013] Furthermore, the extracted objects that can be associated with specific features include: monitoring areas, drones, subtasks, manual executors, ground collaborative vehicles, and collaborative tasks.

[0014] Furthermore, objects have specific characteristics associated with them as follows:

[0015] Specific features associated with monitoring areas include: spatial data of the monitoring area;

[0016] Specific characteristics associated with the drone include: the drone’s operating parameters and operating status;

[0017] The specific features associated with a subtask include: the title of the subtask, the task objective of the subtask, the priority of the subtask, and the necessary information for implementing the subtask;

[0018] Specific features associated with a human performer include: the location and work area of ​​the human performer;

[0019] Specific features associated with the ground cooperative vehicle include: the route and estimated arrival time of the ground cooperative vehicle;

[0020] The specific features associated with collaborative tasks include: the type of collaborative task, the regional scope of the collaborative task, and the priority of the collaborative task.

[0021] Further, the correlations related to the monitoring task include:

[0022] Subtasks are performed in the monitoring area;

[0023] The drone may perform sub-missions;

[0024] Human executors perform collaborative tasks;

[0025] Collaborative ground vehicles perform collaborative tasks;

[0026] Drones may perform collaborative missions;

[0027] Collaborative tasks are grouped into subtasks.

[0028] Furthermore, the flight control strategy includes: flight trajectory, action, speed, acceleration, heading angle, altitude, and photo taking action.

[0029] Furthermore, the flight control model includes a first data structure recognition layer, a second data structure recognition layer, a random factor mixing layer, a tensor layer and a flight control output layer, wherein the first data structure recognition layer inputs structured data and outputs the first hidden features of all objects to the random factor mixing layer, the second data structure recognition layer inputs structured data and outputs the second hidden features of all objects to the random factor mixing layer, the random factor mixing layer introduces randomness and then outputs the third hidden features to the flight control output layer, and the flight control output layer outputs the flight control strategy of the UAV for performing monitoring tasks.

[0030] Furthermore, the first data structure identification layer is a multi-layer structure, and the calculation formula of the first layer is as follows:

[0031]

[0032] represents the first object recognition feature of the vth object in the lth layer, represents the first object recognition feature of the u-th object in the l-1th layer, and Respectively represent the set of objects that have object relationships with object v and object u, represents the cardinality of a set, represents the plane recognition weight matrix of the first data structure recognition layer, , E represents the total number of layers, when l=1 , Represents the features associated with the u-th object, when l=E is equal to the first hidden feature of object v, is the sigmoid function;

[0033] The calculation formula of the second data structure identification layer is as follows: The second data structure identification layer is a multi-layer structure, and the calculation formula of the first layer is as follows:

[0034]

[0035]

[0036] in represents the aggregate representation of the vth object at the lth layer, and denote the second object recognition features of the vth object at the lth layer and the l-1th layer, respectively, Represents a collection of objects that have object relationships with object v; Represents the universal quantifier symbol; represents the second object recognition feature of the u-th object in the l-1th layer, , E represents the total number of layers, when l=1 , Represents the features associated with the u-th object, when l=E is equal to the second hidden feature of object v, represents the aggregation function at level l, represents the plane recognition weight matrix of the lth layer of the recognition layer of the second data structure, σ represents the Sigmoid function;

[0037] The calculation formula of the random factor mixing layer is as follows:

[0038]

[0039] in Represents the third feature matrix, a row vector of which represents the third hidden feature of an object. Represents the first feature matrix, a row vector of which represents the first hidden feature of an object. Represents the second feature matrix, a row vector of which represents the second hidden feature of an object. represents multidimensional noise, represents a multidimensional standard Gaussian distribution, represents the identity matrix;

[0040] The calculation formula of the flight control output layer is as follows:

[0041]

[0042] in represents the first output vector, one component of which represents the probability value of a flight control strategy being selected, and the flight control strategy with the largest probability value is selected as the flight control strategy for the UAV to perform the monitoring task. Represents the output layer weight parameters.

[0043] Furthermore, the land resource and environment monitoring and control system also performs the following: updating structured data according to the image information actually collected by the drone, and the update includes:

[0044] Adding images collected by drones in the monitoring area to specific features associated with the monitoring area;

[0045] The specific features associated with the drone include the images collected by the drone and the actual flight trajectory of the drone;

[0046] The images collected by the drone associated with the subtask are added to the specific features associated with the subtask;

[0047] Then, the aforementioned flight control model is modified, and a branch for identifying changes that affect the monitoring task is added on its basis. The branch includes a task dynamic identification layer, a feature fusion layer, and a task dynamic output layer. The task dynamic identification layer inputs the first feature matrix, outputs the fourth hidden feature matrix to the feature fusion layer and the task dynamic output layer, and the feature fusion layer also inputs the third feature matrix and then outputs the fused features to the flight control output layer. The task dynamic output layer outputs the task objectives of the subtask.

[0048] The calculation formula of the task dynamic identification layer is as follows:

[0049]

[0050] Where D represents the fourth hidden feature matrix, and a row vector thereof represents the fourth hidden feature of an object. represents the first weight parameter, represents the first bias parameter;

[0051] The calculation formula of the feature fusion layer is as follows:

[0052]

[0053] in, represents the fusion feature, represents the feature combination function (concatenation function or summation function), and represent the second and third weight parameters respectively.

[0054] The calculation formula of the modified flight control output layer is as follows:

[0055]

[0056] in represents the second output vector, one component of which represents the probability value of a flight control strategy being selected, and the flight control strategy with the largest probability value is selected as the flight control strategy for the UAV to perform the monitoring task. represents the fourth weight parameter, Represents the second bias parameter.

[0057] The calculation formula of the task dynamic output layer is as follows:

[0058]

[0059] in represents the i-th subtask target vector, whose i-th component represents the probability value of the i-th task target, and the task target with the largest probability value is selected as the current task target of the i-th subtask. represents the fourth hidden feature of the object corresponding to the i-th subtask, represents the fifth weight parameter, Represents the third bias parameter.

[0060] Furthermore, the land resource and environment monitoring and control system also includes a drone control module for controlling the drone to operate according to a flight control strategy through wireless communication.

[0061] The present invention provides a storage medium for storing computer-readable instructions, which can run the aforementioned land resource environment monitoring and control system when read by a computer.

[0062] The beneficial effects of the present invention are:

[0063] The present invention is suitable for complex monitoring tasks of changes in land resources and environment caused by urban expansion and infrastructure construction to automatically generate a UAV flight plan that optimizes the monitoring effect.

[0064] Furthermore, in some embodiments of the present invention, the delay in environmental information collection is taken into account, the feature information is dynamically updated during the strategy execution process, the dynamic changes of subtasks are identified through pattern learning, and adaptive dynamic planning is performed to achieve dynamic adaptation to changes in monitoring requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a module schematic diagram of the land resource environment monitoring and control system of the present invention. DETAILED DESCRIPTION

[0066] The subject matter described herein will now be discussed with reference to example implementations. It should be understood that the discussion of these implementations is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the present specification. Various examples may omit, replace, or add various processes or components as needed. In addition, the features described in some examples may also be combined in other examples.

[0067] At least one embodiment of the present invention discloses a land resource environment monitoring and control system, such as Figure 1 As shown, including:

[0068] The monitoring area spatial data acquisition module 101 is used to acquire spatial data of the monitoring area.

[0069] Spatial data includes geospatial data.

[0070] The UAV data collection module 102 is used to collect the operating parameters and operating status of the UAV;

[0071] The monitoring task parsing module 103 is used to parse the monitoring task to obtain a plurality of operable subtasks.

[0072] For example, the monitoring task of "monitoring changes in land use for expansion and infrastructure projects in a certain area" can be decomposed into subtasks such as "monitoring the expansion area of ​​area A" and "monitoring the infrastructure construction site of area B".

[0073] Of course, monitoring tasks and subtasks do not only include the simple title descriptions mentioned above, but also include necessary information for implementing the tasks and related information such as task objectives. For example, the subtask "Monitoring infrastructure sites in area B" includes the construction information, construction progress, planning information, etc. of the infrastructure sites.

[0074] Collaborative detection information collection module, which is used to collect information that the drone needs to collaborate with the staff;

[0075] Information about collaborative work includes:

[0076] Collaboration object information: location and work area of ​​the manual executor, driving route and estimated arrival time of the ground collaborative vehicle, etc.

[0077] Collaborative task information: the type of monitoring task that needs to be completed in collaboration with other executors, the regional scope and priority of the collaborative task.

[0078] Collaboration time constraints: collaboration window (e.g., when to meet with other executors) and the time urgency of the collaborative task.

[0079] Collaborative communication information: communication channels and methods with other executors, and requirements for sending monitoring data to collaborative objects.

[0080] Collaborative operation mode: Is it parallel operation or does it require drone following / guidance? Is it one-way operation collaboration or two-way interactive collaboration?

[0081] The data structuring module 104 is used to construct structured data including a structural dimension. The basic structural elements included in this structural dimension are objects. The structuring method includes:

[0082] Extract objects that can be associated with specific features;

[0083] Establishing a mapping between the extracted objects and their associated specific features;

[0084] For any two objects, if there is a correlation between the associated specific features related to the monitoring task or there is a correlation between the objects related to the monitoring task, an object relationship is established between the two objects;

[0085] In some embodiments of the present invention, the extracted objects capable of associating specific features include: monitoring areas, drones, subtasks, manual performers, ground collaborative vehicles, and collaborative tasks;

[0086] In some embodiments of the invention, objects have specific features associated with them as follows:

[0087] Specific features associated with monitoring areas include: spatial data of the monitoring area;

[0088] Specific characteristics associated with the drone include: the drone’s operating parameters and operating status;

[0089] The specific features associated with a subtask include: the title of the subtask, the task objective of the subtask, the priority of the subtask, and the necessary information for implementing the subtask;

[0090] Specific features associated with a human performer include: the location and work area of ​​the human performer;

[0091] Specific features associated with the ground cooperative vehicle include: the route and estimated arrival time of the ground cooperative vehicle;

[0092] The specific features associated with collaborative tasks include: the type of collaborative task, the regional scope of the collaborative task, and the priority of the collaborative task.

[0093] In some embodiments of the present invention, the associations related to the monitoring task include:

[0094] Subtasks are performed in the monitoring area;

[0095] The drone may perform sub-missions;

[0096] Human executors perform collaborative tasks;

[0097] Collaborative ground vehicles perform collaborative tasks;

[0098] Drones may perform collaborative missions;

[0099] Collaborative tasks are grouped into subtasks.

[0100] The pattern recognition module 105 is used to perform feature recognition on the structured data through the flight control model to obtain the flight control strategy of all UAVs for performing the monitoring task. The flight control strategy includes: flight trajectory and flight action;

[0101] In some embodiments of the present invention, continuous speed, acceleration, heading angle, etc. are used to represent the action.

[0102] Introducing the altitude dimension allows the drone to adjust its flight altitude.

[0103] Introduce more drone control operations, such as hovering, rotating, taking pictures, etc.

[0104] In some embodiments of the present invention, a drone control module is also included for controlling the drone to operate according to a flight control strategy through wireless communication.

[0105] In some embodiments of the present invention, the flight control model includes a first data structure recognition layer, a second data structure recognition layer, a random factor mixing layer, a tensor layer and a flight control output layer, wherein the first data structure recognition layer inputs structured data and outputs the first hidden features of all objects to the random factor mixing layer, the second data structure recognition layer inputs structured data and outputs the second hidden features of all objects to the random factor mixing layer, the random factor mixing layer introduces randomness and then outputs the third hidden features to the flight control output layer, and the flight control output layer outputs the flight control strategy of the UAV for performing monitoring tasks.

[0106] In some embodiments of the present invention, the flight control output layer outputs a flight control strategy for a UAV to perform a monitoring mission each time.

[0107] In some embodiments of the present invention, the flight control output layer outputs the flight control strategies for all drones to perform monitoring tasks at one time.

[0108] In some embodiments of the present invention, the first data structure identification layer is a multi-layer structure, and the calculation formula of the first layer is as follows:

[0109]

[0110] represents the first object recognition feature of the vth object in the lth layer, represents the first object recognition feature of the u-th object in the l-1th layer, and Respectively represent the set of objects that have object relationships with object v and object u, represents the cardinality of a set, represents the plane recognition weight matrix of the first data structure recognition layer, , E represents the total number of layers, when l=1 , Represents the features associated with the u-th object, when l=E is equal to the first hidden feature of object v, is the sigmoid function;

[0111] The calculation formula of the second data structure identification layer is as follows: The second data structure identification layer is a multi-layer structure, and the calculation formula of the first layer is as follows:

[0112]

[0113]

[0114] in represents the aggregate representation of the vth object at the lth layer, and denote the second object recognition features of the vth object at the lth layer and the l-1th layer, respectively, Represents a collection of objects that have object relationships with object v; Represents the universal quantifier symbol; represents the second object recognition feature of the u-th object in the l-1th layer, , E represents the total number of layers (since the input of the second data structure recognition layer is similar to that of the first data structure recognition layer, the number of layers is the same, both E), when l=1 , Represents the features associated with the u-th object, when l=E is equal to the second hidden feature of object v, represents the aggregation function of layer l (average or maximum pooling function), represents the plane recognition weight matrix of the lth layer of the second data structure recognition layer, and σ represents the Sigmoid function.

[0115] The calculation formula of the random factor mixing layer is as follows:

[0116]

[0117] in Represents the third feature matrix, a row vector of which represents the third hidden feature of an object. Represents the first feature matrix, a row vector of which represents the first hidden feature of an object. Represents the second feature matrix, a row vector of which represents the second hidden feature of an object. represents multidimensional noise, represents a multidimensional standard Gaussian distribution, Represents the identity matrix.

[0118] The calculation formula of the flight control output layer is as follows:

[0119]

[0120] in represents the first output vector, one component of which represents the probability value of a flight control strategy being selected, and the flight control strategy with the largest probability value is selected as the flight control strategy for the UAV to perform the monitoring task. Represents the output layer weight parameters.

[0121] In some further embodiments of the present invention, due to the presence of dynamic human activities such as urban expansion and infrastructure construction in the monitoring area, the impact of these activities on the environment of the monitoring area is also dynamically changing, the collected spatial data has a certain lag, and urban expansion and infrastructure construction will also cause the predetermined monitoring tasks to no longer adapt to the changed area.

[0122] First, the structured data is updated based on the image information actually collected by the drone. The update includes:

[0123] Adding images collected by drones in the monitoring area to specific features associated with the monitoring area;

[0124] The specific features associated with the drone include the images collected by the drone and the actual flight trajectory of the drone;

[0125] The images collected by the drone associated with the subtask are added to the specific features associated with the subtask;

[0126] Then, the aforementioned flight control model is modified, and a branch for identifying changes that affect the monitoring task is added on its basis. The branch includes a task dynamic identification layer, a feature fusion layer, and a task dynamic output layer. The task dynamic identification layer inputs the first feature matrix, outputs the fourth hidden feature matrix to the feature fusion layer and the task dynamic output layer, and the feature fusion layer also inputs the third feature matrix and then outputs the fused features to the flight control output layer. The task dynamic output layer outputs the task objectives of the subtask.

[0127] The calculation formula of the task dynamic identification layer is as follows:

[0128]

[0129] Where D represents the fourth hidden feature matrix, and a row vector thereof represents the fourth hidden feature of an object. represents the first weight parameter, represents the first bias parameter;

[0130] The calculation formula of the feature fusion layer is as follows:

[0131]

[0132] in, represents the fusion feature, represents the feature combination function (concatenation function or summation function), and represent the second and third weight parameters respectively.

[0133] The calculation formula of the modified flight control output layer is as follows:

[0134]

[0135] in represents the second output vector, one component of which represents the probability value of a flight control strategy being selected, and the flight control strategy with the largest probability value is selected as the flight control strategy for the UAV to perform the monitoring task. represents the fourth weight parameter, Represents the second bias parameter.

[0136] The calculation formula of the task dynamic output layer is as follows:

[0137]

[0138] in represents the i-th subtask target vector, whose i-th component represents the probability value of the i-th task target, and the task target with the largest probability value is selected as the current task target of the i-th subtask. represents the fourth hidden feature of the object corresponding to the i-th subtask, represents the fifth weight parameter, Represents the third bias parameter.

[0139] The task target output by the task dynamic output layer comes from a task target library, which contains multiple executable task targets that can be used as task targets of subtasks, not the collected task targets.

[0140] The branch used to identify changes affected by the monitoring task is independently pre-trained. During the pre-training, the parameters are updated by back-propagation through the output loss of the task dynamic output layer, so that it has the ability to dynamically judge the changes in the task objectives of the original sub-task based on the information collected by the UAV, and the fourth hidden feature attached to the information is input into the flight control output layer as the identification basis for obtaining the flight control strategy used by the UAV to perform the monitoring task.

[0141] In some embodiments of the present invention, the task objectives include monitoring the bare soil area, monitoring the soil erosion area, etc.

[0142] The above describes an embodiment of the present invention, but this embodiment is not limited to the above-mentioned specific implementation mode. The above-mentioned specific implementation mode is merely illustrative and not restrictive. Under the guidance of this embodiment, ordinary technicians in this field can also make more forms of equivalent embodiments, all of which are within the protection of this embodiment.

Claims

1. A land resource environment monitoring and control system, It is characterized in that include: A monitoring area spatial data acquisition module, which is used to collect spatial data of the monitoring area; UAV data collection module, which is used to collect the operating parameters and operating status of the UAV; A monitoring task parsing module, which is used to parse the monitoring task to obtain multiple operable subtasks; The data structuring module is used to construct structured data including a structural dimension. The basic structural elements contained in this structural dimension are objects. The structuring methods include: Extract objects that can be associated with specific features; Establishing a mapping between the extracted objects and their associated specific features; For any two objects, if there is a correlation between the associated specific features related to the monitoring task or there is a correlation between the objects related to the monitoring task, an object relationship is established between the two objects; The pattern recognition module is used to perform feature recognition on structured data through the flight control model to obtain the flight control strategy used by all drones to perform monitoring tasks.

2. A land resource environment monitoring and control system according to claim 1, It is characterized in that The extracted objects that can be associated with specific features include: monitoring area, drone, subtask, manual executor, ground collaborative vehicle, and collaborative task.

3. A land resource environment monitoring and control system according to claim 2, It is characterized in that Objects have specific characteristics associated with them: Specific features associated with monitoring areas include: spatial data of the monitoring area; Specific characteristics associated with the drone include: the drone’s operating parameters and operating status; The specific features associated with a subtask include: the title of the subtask, the task objective of the subtask, the priority of the subtask, and the necessary information for implementing the subtask; Specific features associated with a human performer include: the location and work area of ​​the human performer; Specific features associated with the ground cooperative vehicle include: the route and estimated arrival time of the ground cooperative vehicle; The specific features associated with collaborative tasks include: the type of collaborative task, the regional scope of the collaborative task, and the priority of the collaborative task.

4. A land resource environment monitoring and control system according to claim 3, It is characterized in that Relevance to monitoring tasks include: Subtasks are performed in the monitoring area; The drone may perform sub-missions; Human executors perform collaborative tasks; Collaborative ground vehicles perform collaborative tasks; Drones may perform collaborative missions; Collaborative tasks are grouped into subtasks.

5. A land resource environment monitoring and control system according to claim 1, It is characterized in that The flight control strategy includes: flight trajectory, action, speed, acceleration, heading angle, altitude, and photo taking action.

6. A land resource environment monitoring and control system according to claim 1, It is characterized in that The flight control model includes a first data structure recognition layer, a second data structure recognition layer, a random factor mixing layer, a tensor layer and a flight control output layer, wherein the first data structure recognition layer inputs structured data and outputs the first hidden features of all objects to the random factor mixing layer, the second data structure recognition layer inputs structured data and outputs the second hidden features of all objects to the random factor mixing layer, the random factor mixing layer introduces randomness and then outputs the third hidden features to the flight control output layer, and the flight control output layer outputs the flight control strategy used by the UAV to perform monitoring tasks.

7. A land resource environment monitoring and control system according to claim 6, It is characterized in that The first data structure recognition layer is a multi-layer structure, and the calculation formula of the first layer is as follows: ； represents the first object recognition feature of the vth object in the lth layer, represents the first object recognition feature of the u-th object in the l-1th layer, and Respectively represent the set of objects that have object relationships with object v and object u, represents the cardinality of a set, represents the plane recognition weight matrix of the first data structure recognition layer, , E represents the total number of layers, when l=1 , Represents the features associated with the u-th object, when l=E is equal to the first hidden feature of object v, is the sigmoid function; The calculation formula of the second data structure identification layer is as follows: The second data structure identification layer is a multi-layer structure, and the calculation formula of the first layer is as follows: ； ； in represents the aggregate representation of the vth object at the lth layer, and denote the second object recognition features of the vth object at the lth layer and the l-1th layer, respectively, Represents a collection of objects that have object relationships with object v; Represents the universal quantifier symbol; represents the second object recognition feature of the u-th object in the l-1th layer, , E represents the total number of layers, when l=1 , Represents the feature associated with the u-th object, when l=E is equal to the second hidden feature of object v, represents the aggregation function at level l, represents the plane recognition weight matrix of the lth layer of the recognition layer of the second data structure, σ represents the Sigmoid function; The calculation formula of the random factor mixing layer is as follows: ； in Represents the third feature matrix, a row vector of which represents the third hidden feature of an object. Represents the first feature matrix, a row vector of which represents the first hidden feature of an object. Represents the second feature matrix, a row vector of which represents the second hidden feature of an object. represents multidimensional noise, represents a multidimensional standard Gaussian distribution, represents the identity matrix; The calculation formula of the flight control output layer is as follows: ； in represents the first output vector, one component of which represents the probability value of a flight control strategy being selected, and the flight control strategy with the largest probability value is selected as the flight control strategy for the UAV to perform the monitoring mission. Represents the output layer weight parameters.

8. A land resource environment monitoring and control system according to claim 7, It is characterized in that Update structured data based on the image information actually collected by the drone. The update includes: Adding images collected by drones in the monitoring area to specific features associated with the monitoring area; The specific features associated with the drone include the images collected by the drone and the actual flight trajectory of the drone; The images collected by the drone associated with the subtask are added to the specific features associated with the subtask; Then, the aforementioned flight control model is modified, and a branch for identifying changes that affect the monitoring task is added on its basis, the branch including a task dynamic identification layer, a feature fusion layer and a task dynamic output layer, wherein the task dynamic identification layer inputs the first feature matrix, outputs the fourth hidden feature matrix to the feature fusion layer and the task dynamic output layer, the feature fusion layer also inputs the third feature matrix, and then outputs the fused feature to the flight control output layer, and the task dynamic output layer outputs the task objective of the subtask; The calculation formula of the task dynamic identification layer is as follows: ； Where D represents the fourth hidden feature matrix, and a row vector thereof represents the fourth hidden feature of an object. represents the first weight parameter, represents the first bias parameter; The calculation formula of the feature fusion layer is as follows: ； in, represents the fusion feature, represents the feature combination function (concatenation function or summation function), and denote the second and third weight parameters respectively; The calculation formula of the modified flight control output layer is as follows: ； in represents the second output vector, one component of which represents the probability value of a flight control strategy being selected, and the flight control strategy with the largest probability value is selected as the flight control strategy for the UAV to perform the monitoring task. represents the fourth weight parameter, represents the second bias parameter; The calculation formula of the task dynamic output layer is as follows: ； in represents the i-th subtask target vector, whose i-th component represents the probability value of the i-th task target, and the task target with the largest probability value is selected as the current task target of the i-th subtask. represents the fourth hidden feature of the object corresponding to the i-th subtask, represents the fifth weight parameter, Represents the third bias parameter.

9. A land resource environment monitoring and control system according to claim 1, It is characterized in that It also includes a drone control module for controlling the drone to operate according to a flight control strategy through wireless communication.

10. A storage medium, It is characterized in that It is used to store computer-readable instructions, which, when read by a computer, can run a land resource environment monitoring and control system as described in any one of claims 1-9.

Images