Data information collaborative management method based on deep learning
Through the deep learning-based data information collaborative management method, the problems of task allocation and accurate trajectory positioning in the collaboration process are solved, and efficient collaborative management and safe flight between drones are realized.
Patent Information
- Application Number
- CN202411819076.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-11
- Publication Date
- 2025-03-25
AI Technical Summary
It is difficult for UAVs to allocate tasks during joint collaboration, and the lack of effective work coordination and accurate trajectory positioning leads to problems such as yaw and collision.
The data information collaborative management method based on deep learning is adopted. By obtaining the information of multiple drones, the corresponding relationship between the behavior information and the drone end is established, the flight model and rehearsal model are constructed, and the drone trajectory is monitored and adjusted in real time to ensure the completion and safety of the task.
It realizes efficient collaborative management between drones, avoids collision and yaw, and improves work management efficiency and accuracy.
Smart Images
Figure CN119694170A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of information collaboration technology, and in particular to a data information collaborative management method based on deep learning. Background Art
[0002] With the continuous advancement and popularization of drone technology, more and more drone manufacturers and service providers have appeared in the market. This competition and innovation situation has promoted the technical innovation and application expansion of drone data information collaborative management, and also made various manufacturers explore the market suitable for themselves. In the process of joint collaboration, it is difficult to assign tasks to drones, there is no good work coordination, and it is difficult to accurately locate the trajectory of drones and confirm the yaw. Summary of the invention
[0003] The purpose of the present invention is to provide a data information collaborative management method based on deep learning to address the deficiencies in the background technology.
[0004] In order to achieve the above object, the present invention provides the following technical solution: a data information collaborative management method based on deep learning, comprising the following steps:
[0005] Obtain information of multiple drones, where the drone information includes basic drone information and drone type information, and obtain multiple drone terminals based on the registration of multiple drone information;
[0006] Set a drone type management area based on drone type information, and classify and manage drone terminals based on the drone type management area;
[0007] Corresponding behavior information is formulated according to the drone type information, wherein the behavior information includes multiple functional information of the drone, and a corresponding relationship between the behavior information and the drone end is established based on deep learning to obtain a learning model;
[0008] Acquire the work site information of the UAV, formulate the UAV flight information based on the work site information, and construct a flight model with the work site information and the UAV flight information, wherein the UAV flight information includes the UAV flight route information and the corresponding flight mission information;
[0009] Extract the behavior information from the flight mission information, obtain the UAV end that matches the behavior information based on the learning model as the mission UAV end, and build a rehearsal model based on the mission UAV end and the flight model;
[0010] Establish the corresponding information between the rehearsal model and the real-time flight model. The mission UAV performs flight based on the UAV flight information. The trajectory of the mission UAV is obtained and transmitted to the flight model in real time to obtain the real-time flight model.
[0011] The real-time flight model corresponding to the corresponding information that does not meet the preset conditions is used as the impact model, and the rehearsal model is re-rehearsed based on the impact model until all the mission drone ends complete the drone flight information.
[0012] In a preferred embodiment, the step of classifying and managing the drone terminals based on the drone type management area includes:
[0013] Set the drone type management area based on the drone type information, and bind the drone type information with the corresponding drone type management area;
[0014] Based on the drone information, the drone type information corresponding to the drone end is obtained, and the drone end is matched with the drone type information and stored in the corresponding drone type management area.
[0015] In a preferred embodiment, the step of establishing a corresponding relationship between behavior information and the drone end based on deep learning to obtain a learning model includes:
[0016] Formulate corresponding behavior information according to the drone type information, and bind the behavior information with the drone type information;
[0017] Fusing the behavior information with the drone type information to obtain fused data;
[0018] The fused data is divided into training set and test set, the training set is input into the neural network model, and the updated neural network model is obtained through multiple rounds of iterations. The test set is input into the neural network model for verification to obtain the learning model.
[0019] In a preferred embodiment, the step of constructing a flight model using the work site information and the UAV flight information includes:
[0020] Obtaining the work site information of the drone, wherein the work site information includes the geographical information of the work environment and the work range information;
[0021] A site model is constructed based on the work site information, and the site model is divided into types to obtain multiple model sections and marked in the site model;
[0022] Plan the UAV flight route information and corresponding flight mission information based on the site model and use it as UAV flight information;
[0023] The UAV flight information is constructed in the site model to obtain a flight model, multiple marking points are selected in the UAV flight route, and the corresponding information between the site model and the marking points is extracted, wherein the corresponding information includes the angle information and distance information between the lower projection point position of the marking point in the site model and the adjacent marking points.
[0024] In a preferred embodiment, the step of planning the UAV flight route information and the corresponding flight mission information based on the site model and using the UAV flight information as the UAV flight information includes:
[0025] Obtaining an upper projection surface of the site model, and marking the upper projection surface of the model plate on the upper projection surface of the site model to obtain a first reference surface;
[0026] Based on the first reference surface, the model plate with the same plate task information is set as a flight range surface, and the horizontal flight route information is planned in the flight range surface;
[0027] Acquire the altitude information of the site model corresponding to the projection under the horizontal flight route information as the altitude information;
[0028] Formulate a preset flight altitude for the drone, and add the preset flight altitude of the drone to the altitude information to obtain vertical flight route information;
[0029] The vertical flight route information and the horizontal flight route information are used to construct a three-dimensional flight route as the UAV flight route;
[0030] The section mission information is used as the flight mission information, and the UAV flight route information and the corresponding flight mission information are used as the UAV flight information.
[0031] In a preferred embodiment, the steps of extracting behavior information from the flight mission information, obtaining a UAV end matching the behavior information based on the learning model as the mission UAV end, and building a rehearsal model based on the mission UAV end and the flight model include:
[0032] Extract the behavior information from the flight mission information and input it into the learning model, and output the corresponding drone end through the learning model as the mission drone end;
[0033] The mission UAV is rehearsed according to the UAV flight information in the flight model to obtain a rehearsal model.
[0034] In a preferred embodiment, the step of establishing corresponding information between the preview model and the real-time flight model, the task UAV terminal flying based on the UAV flight information, obtaining the trajectory of the task UAV terminal and transmitting it to the flight model in real time to obtain the real-time flight model includes:
[0035] Obtaining the rehearsal time information corresponding to the UAV flight route information in the flight mission information in the rehearsal model, and matching the UAV flight route information in the rehearsal model with the identification points in the UAV flight route information of the flight model based on the rehearsal time information to obtain corresponding information;
[0036] The task UAV is flown in the work area according to the UAV flight information and the real-time trajectory of the task UAV is obtained, and the real-time trajectory is constructed in the flight model to obtain a real-time flight model;
[0037] The trajectory accuracy index is obtained based on the coincidence information between the real-time flight model and the preview model.
[0038] In a preferred embodiment, the step of using the real-time flight model corresponding to the trajectory accuracy index that does not meet the preset conditions as the impact model, and re-rehearsing the rehearsal model based on the impact model until all the mission drone terminals complete the drone flight information includes:
[0039] Set preset conditions, and use the real-time flight model corresponding to the trajectory accuracy index that does not meet the preset conditions as the influencing model;
[0040] Based on the impact model, the response rehearsal model is re-rehearsed to obtain an updated rehearsal model, and whether there is conflict information on the task UAV side in the updated rehearsal model, wherein the conflict information is a collision situation between the task UAV side and the UAV flight route information;
[0041] If there is conflicting information on the mission drone side in the updated rehearsal model, the mission drone side will be adjusted in flight trajectory or flight speed until all mission drone sides complete the drone flight information.
[0042] In the above technical solution, the technical effects and advantages provided by the present invention are:
[0043] The present invention can obtain a trajectory accuracy index according to the overlap information between the real-time flight model and the preview model, has a good collaborative management effect between UAVs, avoids collisions between subsequent UAVs, and has a good work management effect. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.
[0045] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION
[0046] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0047] Example 1, please refer to Figure 1 As shown, the data information collaborative management method based on deep learning described in this embodiment includes the following steps:
[0048] S1. Acquire information of multiple drones, wherein the drone information includes basic drone information and drone type information, and obtain multiple drone terminals based on the registration of multiple drone information;
[0049] S2. Setting a drone type management area based on drone type information, and classifying and managing drone terminals based on the drone type management area;
[0050] S3. Formulate corresponding behavior information according to the drone type information, wherein the behavior information includes multiple functional information of the drone, and establish a corresponding relationship between the behavior information and the drone end based on deep learning to obtain a learning model;
[0051] S4, obtaining the work site information of the UAV, formulating the UAV flight information based on the work site information, and constructing a flight model by combining the work site information and the UAV flight information, wherein the UAV flight information includes the UAV flight route information and the corresponding flight mission information;
[0052] S5, extracting behavior information from the flight mission information, obtaining a UAV end that matches the behavior information based on the learning model as the mission UAV end, and building a rehearsal model based on the mission UAV end and the flight model;
[0053] S6. Establishing corresponding information between the rehearsal model and the real-time flight model, the mission UAV performs flight based on the UAV flight information, and obtains the trajectory of the mission UAV and transmits it to the flight model in real time to obtain the real-time flight model;
[0054] S7, taking the real-time flight model corresponding to the corresponding information that does not meet the preset conditions as the impact model, and re-rehearsing the response rehearsal model based on the impact model until all the mission drones complete the drone flight information;
[0055] As described in the above steps S1-S7, with the continuous advancement and popularization of drone technology, more and more drone manufacturers and service providers have appeared in the market. This situation of competition and innovation has promoted the technological innovation and application expansion of drone data information collaborative management, and has also made various manufacturers explore markets suitable for themselves. In the process of joint collaboration, it is difficult for drones to assign tasks to drones, and there is no good work coordination. It is difficult to accurately locate the trajectory of the drone and confirm the yaw. The present application can obtain the trajectory accuracy index based on the overlap information between the real-time flight model and the preview model, and has a good collaborative management effect between drones, avoiding collisions between subsequent drones, and has a good work management effect;
[0056] In one embodiment, the step S2 of classifying and managing the drone terminals based on the drone type management area includes:
[0057] S21. Setting a drone type management area based on the drone type information, and binding the drone type information with the corresponding drone type management area;
[0058] S22, based on the drone information, obtain drone type information corresponding to the drone end, and store the drone type management area corresponding to the drone end and the drone type information;
[0059] As described in the above steps S21 and S22, it is first necessary to obtain information on multiple drones. The drone information here includes basic drone information and drone type information, wherein the basic drone information is the performance parameter information of the drone, and the drone type information is the working function of the drone. For example, the drone can be responsible for collecting parameters such as shooting and environment. There are multiple types of drones. Based on the registration of multiple drone information, multiple drone terminals are obtained, and then multiple mission machine type management areas can be set. The drone type management area is an independent storage space. The drone type information is bound to the corresponding drone type management area, then different drone type management areas are used to store different types of drones. When a drone When the human-machine has multiple drone type information, then the drone can be stored in multiple corresponding drone type management areas, and the drone end and the drone type information are matched to the corresponding drone type management area for corresponding storage. Here, the drone end and the corresponding drone type management area are in an information binding state. When a drone end is stored in multiple drone type management areas, if the drone end is subsequently selected as a task drone end, the drone end will be removed from the drone type management area, indicating that it can no longer be selected. After the task drone end is used, it will be classified into the corresponding drone type management area again, which has a good management effect on drones and can greatly improve the dispatch of drones;
[0060] In one embodiment, the step S3 of establishing the correspondence between the behavior information and the drone end based on deep learning to obtain a learning model includes:
[0061] S31. Formulate corresponding behavior information corresponding to the drone type information, and bind the behavior information with the drone type information;
[0062] S32, fusing the behavior information with the drone type information to obtain fused data;
[0063] S33, dividing the fused data into a training set and a test set, inputting the training set into the neural network model, obtaining an updated neural network model through multiple rounds of iterations, and inputting the test set into the neural network model for verification to obtain a learning model;
[0064] As described in the above steps S31-S33, it is necessary to formulate corresponding behavior information corresponding to the drone type information, wherein the behavior information includes multiple functional information of the drone. For example, the drone has multifunctional drones such as shooting, temperature monitoring, wind speed monitoring, etc. Here, the behavior information can be: shooting, temperature monitoring, wind speed monitoring and other functions, the drone type information and the corresponding behavior information are bound, and then the neural network model is trained, and the behavior information is input into the neural network model to obtain the corresponding drone type information. Here, it is used as a training data set, and then the neural network model can be trained to obtain a learning model. Then, by inputting the required behavior information, the corresponding drone type information can be obtained through the learning model, which can greatly improve the efficiency of drone selection, have a good drone information synergy effect, and ensure efficient drone selection;
[0065] In one embodiment, the step S4 of constructing a flight model using the work site information and the UAV flight information includes:
[0066] S41, obtaining the work site information of the drone, wherein the work site information includes the geographical information of the work environment and the work range information;
[0067] S42, constructing a site model based on the work site information, dividing the site model into types to obtain multiple model sections and marking them in the site model;
[0068] S43, planning the UAV flight route information and the corresponding flight mission information based on the site model and using them as the UAV flight information;
[0069] S44, constructing the UAV flight information in the site model to obtain a flight model, selecting multiple marking points in the UAV flight route, and extracting corresponding information between the site model and the marking points, wherein the corresponding information includes angle information and distance information between the lower projection point position of the marking point in the site model and the adjacent marking points;
[0070] As described in the above steps S41-S44, the UAV needs a work location when working, so the work site information needs to be collected before work, and then a site model is constructed based on the collected work site information. The scene model is a geographical three-dimensional model within the scope of the work range information, and then the site model is divided into types to obtain multiple model sections. For example, there will be different geographical environments in the scene model, such as lakes, plains or mountains and other different scene ranges, which are divided here as multiple model sections and marked in the site model. Then, the UAV flight route information and the corresponding flight mission information are planned based on the site model and used as the UAV flight information. The site model here can plan the route according to the model section when planning the UAV flight route information. For example, continuous route planning can be performed in adjacent site models of the same type, so that the UAV can perform tasks more clearly and it is easier to perform subsequent task allocation to the UAV, making the UAV work tasks more clear. The relatively simple flight mission information is obtained by single work to avoid the tediousness and complexity of the UAV flight mission information, and can reduce the diversity of the UAV collected data. This can also facilitate the coordinated management and data processing of the subsequently collected information, ensure the smoothness of the UAV operation, and better coordinate the management of the UAV. After that, after determining the UAV flight information, the UAV flight information is constructed in the site model to obtain the flight model, and multiple marking points are selected in the UAV flight route. Based on the flight model, the corresponding information between the site model and the marking points is extracted, wherein the corresponding information includes the angle information and distance information between the lower projection point position of the marking point in the site model and the adjacent marking points, wherein the marking point is selected to set a marking point at a preset distance in the UAV flight route and to set a marking point at a position with a turning point of the route trajectory in the UAV flight route. The marking point can be used to subsequently verify the accuracy of the UAV flight route to avoid the UAV route deviation affecting the accuracy of the collected data;
[0071] In one embodiment, the step S43 of planning the UAV flight route information and the corresponding flight mission information based on the site model and using the UAV flight information includes:
[0072] S431, obtaining an upper projection surface of the site model, and marking the upper projection surface of the model block on the upper projection surface of the site model to obtain a first reference surface;
[0073] S432, setting the module task information based on the model module corresponding to the first reference surface, taking the model module with the same module task information in the first reference surface as a flight range surface, and planning the horizontal flight route information in the flight range surface;
[0074] S433, obtaining the altitude information of the site model corresponding to the projection under the horizontal flight route information as the altitude information;
[0075] S434, setting a preset flight altitude for the UAV, and adding the preset flight altitude for the UAV to the altitude information to obtain vertical flight route information;
[0076] S435, constructing a three-dimensional flight route using the vertical flight route information and the horizontal flight route information as the UAV flight route;
[0077] S436, using the section mission information as flight mission information, and the UAV flight route information and the corresponding flight mission information as UAV flight information;
[0078] As described in the above steps S431-S436, here we first obtain the upper projection surface of the site model, and mark the upper projection surface of the model panel in the upper projection surface of the site model to obtain the first reference surface, set the panel task information based on the model panel corresponding to the first reference surface, and use the model panel with the same panel task information in the first reference surface as a flight range surface, and plan the horizontal flight route information in the flight range surface. The horizontal flight route information here is a mapped flight route on a horizontal plane, and then obtain the altitude information of the site model corresponding to the projection under the horizontal flight route information as the altitude information. Here, the site model corresponding to the projection under the horizontal flight route information can be obtained, which is used here according to the geographical environment of the flight. Set a reasonable flight altitude, formulate a preset flight altitude of the UAV, add the preset flight altitude of the UAV to the altitude information to obtain the vertical flight route information, where the preset flight altitude of the UAV is a reasonable flight altitude, then add the preset flight altitude of the UAV to the altitude information to obtain the vertical flight route information, and construct a three-dimensional flight route as the UAV flight route by combining the vertical flight route information with the horizontal flight route information. Here, the UAV flight altitude route of the vertical flight route information and the horizontal flight route information are integrated to obtain a three-dimensional UAV flight route as the UAV flight route, which can greatly improve the planning efficiency of the UAV flight route and ensure the rationality of the planned route;
[0079] In one embodiment, the step S5 of extracting behavior information from the flight mission information, obtaining a UAV end matching the behavior information based on the learning model as the mission UAV end, and building a rehearsal model based on the mission UAV end and the flight model includes:
[0080] S51, extracting behavior information from the flight mission information and inputting it into the learning model, and outputting the corresponding UAV end as the mission UAV end through the learning model;
[0081] S52, performing a flight rehearsal on the mission UAV according to the UAV flight information in the flight model to obtain a rehearsal model;
[0082] As described in the above steps S51 and S52, after determining the UAV flight information, it is necessary to select the UAV that performs the task, extract the behavior information in the flight mission information, and then input the behavior information into the learning model and output the corresponding UAV end as the task UAV. Then, the task UAV is removed from the corresponding UAV type management area, indicating that the UAV in the UAV type management area cannot be selected again. Then, the task UAV end is rehearsed according to the UAV flight information in the flight model to obtain a rehearsal model, which can understand the state of the task UAV during normal flight and has a good reference effect.
[0083] In one embodiment, the step S6 of establishing corresponding information between the preview model and the real-time flight model, wherein the task UAV terminal flies based on the UAV flight information, and the trajectory of the task UAV terminal is obtained and transmitted to the flight model in real time to obtain the real-time flight model includes:
[0084] S61, obtaining rehearsal time information corresponding to the UAV flight route information in the flight mission information in the rehearsal model, and matching the UAV flight route information in the rehearsal model with the identification points in the UAV flight route information of the flight model based on the rehearsal time information to obtain corresponding information;
[0085] S62, flying the task UAV in the work site according to the UAV flight information and obtaining the real-time trajectory of the task UAV, and constructing the real-time trajectory into the flight model to obtain a real-time flight model;
[0086] S63, obtaining a trajectory accuracy index according to the overlap information between the real-time flight model and the preview model, wherein the calculation formula of the trajectory accuracy index is:
[0087] ,in, is the trajectory accuracy index, is the position difference between the lower projection point position of the mark point on the real-time trajectory in the site model and the lower projection point position of the mark point in the corresponding information in the site model. It is the angle difference between the lower projection point of the mark point on the real-time track and the adjacent mark point in the site model. It is the distance information difference value between the lower projection point position of the mark point on the real-time track in the site model and one of the adjacent mark points. It is the distance information gap value between the lower projection point position of the mark point on the real-time trajectory in the site model and another adjacent mark point. is the position gap weight coefficient, is the angle difference weight coefficient, and is a constant greater than zero;
[0088] As described in the above steps S61-S63, the rehearsal time information corresponding to the UAV flight route information in the flight mission information in the rehearsal model can be obtained here, and the UAV flight route information in the rehearsal model is matched with the identification points in the UAV flight route information of the flight model based on the rehearsal time information to obtain corresponding information. This is to match the routes of the UAV flight route information in the rehearsal model and the flight model, and then the task UAV end is flown in the work site according to the UAV flight information and the real-time trajectory of the task UAV end is obtained. The real-time trajectory is constructed in the flight model to obtain a real-time flight model, and the trajectory accuracy index can be obtained according to the overlap information between the real-time flight model and the rehearsal model, which has a good collaborative management effect between UAVs, avoids subsequent UAVs from colliding, and has a good work management effect;
[0089] In one embodiment, the step S7 of using the real-time flight model corresponding to the trajectory accuracy index that does not meet the preset conditions as the impact model, and re-rehearsing the rehearsal model based on the impact model until all the mission drones complete the drone flight information includes:
[0090] S71, setting a preset condition, and taking the real-time flight model corresponding to the trajectory accuracy index that does not meet the preset condition as the influencing model;
[0091] S72, re-rehearsing the response rehearsal model based on the impact model to obtain an updated rehearsal model, and obtaining whether there is conflict information on the task UAV side in the updated rehearsal model, wherein the conflict information is a collision situation between the task UAV side and the UAV flight route information;
[0092] S73. If conflicting information exists on the task UAV end in the updated rehearsal model, the flight trajectory or flight speed of the task UAV end is adjusted until all the task UAV ends complete the UAV flight information;
[0093] As described in the above steps S71-S73, preset conditions are set, and the real-time flight model corresponding to the trajectory accuracy index that does not meet the preset conditions is used as the influencing model. The preset conditions here are the settings of the corresponding trajectory accuracy index, which are used to limit the trajectory accuracy index. Then, based on the influencing model, the rehearsal model is re-rehearsed to obtain an updated rehearsal model, and whether there is conflict information on the task drone end in the updated rehearsal model, wherein the conflict information is the collision of the drone flight route information on the task drone end, which can avoid collisions between task drone ends. If there is conflict information on the task drone end in the updated rehearsal model, the flight trajectory or flight speed of the task drone end will be adjusted until all the task drone ends complete the drone flight information, which can ensure that the task drone ends can complete the work more smoothly.
[0094] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A data information collaborative management method based on deep learning, characterized in that: The following steps are involved: Obtain information of multiple drones, where the drone information includes basic drone information and drone type information, and obtain multiple drone terminals based on the registration of multiple drone information; Set a drone type management area based on drone type information, and classify and manage drone terminals based on the drone type management area; Corresponding behavior information is formulated according to the drone type information, wherein the behavior information includes multiple functional information of the drone, and a corresponding relationship between the behavior information and the drone end is established based on deep learning to obtain a learning model; Acquire the work site information of the UAV, formulate the UAV flight information based on the work site information, and construct a flight model with the work site information and the UAV flight information, wherein the UAV flight information includes the UAV flight route information and the corresponding flight mission information; Extract the behavior information from the flight mission information, obtain the UAV end that matches the behavior information based on the learning model as the mission UAV end, and build a rehearsal model based on the mission UAV end and the flight model; Establish the corresponding information between the rehearsal model and the real-time flight model. The mission UAV performs flight based on the UAV flight information. The trajectory of the mission UAV is obtained and transmitted to the flight model in real time to obtain the real-time flight model. The real-time flight model corresponding to the corresponding information that does not meet the preset conditions is used as the impact model, and a rehearsal model corresponding to the impact model is re-rehearsed until all the mission drone terminals complete the drone flight information.
2. The method for collaborative management of data information based on deep learning according to claim 1, characterized in that: The step of classifying and managing the drone terminals based on the drone type management area includes: Set the drone type management area based on the drone type information, and bind the drone type information with the corresponding drone type management area; Based on the drone information, the drone type information corresponding to the drone end is obtained, and the drone end is matched with the drone type information and stored in the corresponding drone type management area.
3. The method for collaborative management of data information based on deep learning according to claim 1, characterized in that: The step of establishing the corresponding relationship between the behavior information and the drone end based on deep learning to obtain a learning model includes: Formulate corresponding behavior information according to the drone type information, and bind the behavior information with the drone type information; Fusing the behavior information with the drone type information to obtain fused data; The fused data is divided into training set and test set, the training set is input into the neural network model, and the updated neural network model is obtained through multiple rounds of iterations. The test set is input into the neural network model for verification to obtain the learning model.
4. The method for collaborative management of data information based on deep learning according to claim 1, characterized in that: The step of constructing a flight model using the work site information and the UAV flight information includes: Obtaining the work site information of the drone, wherein the work site information includes the geographical information of the work environment and the work range information; A site model is constructed based on the work site information, and the site model is divided into types to obtain multiple model sections and marked in the site model; Plan the UAV flight route information and corresponding flight mission information based on the site model and use it as UAV flight information; The UAV flight information is constructed in the site model to obtain a flight model, multiple marking points are selected in the UAV flight route, and the corresponding information between the site model and the marking points is extracted, wherein the corresponding information includes the angle information and distance information between the lower projection point position of the marking point in the site model and the adjacent marking points.
5. The method for collaborative management of data information based on deep learning according to claim 4, characterized in that: The step of planning the UAV flight route information and the corresponding flight mission information based on the site model and using the UAV flight information as the UAV flight information includes: Obtaining an upper projection surface of the site model, and marking the upper projection surface of the model plate on the upper projection surface of the site model to obtain a first reference surface; Based on the first reference surface, the model plate with the same plate task information is set as a flight range surface, and the horizontal flight route information is planned in the flight range surface; Acquire the altitude information of the site model corresponding to the projection under the horizontal flight route information as the altitude information; Formulate a preset flight altitude for the drone, and add the preset flight altitude of the drone to the altitude information to obtain vertical flight route information; The vertical flight route information and the horizontal flight route information are used to construct a three-dimensional flight route as the UAV flight route; The section mission information is used as the flight mission information, and the UAV flight route information and the corresponding flight mission information are used as the UAV flight information.
6. The method for collaborative management of data information based on deep learning according to claim 1, characterized in that: The steps of extracting behavior information from the flight mission information, obtaining a UAV end matching the behavior information based on the learning model as the mission UAV end, and building a rehearsal model based on the mission UAV end and the flight model include: Extract the behavior information from the flight mission information and input it into the learning model, and output the corresponding drone end through the learning model as the mission drone end; The mission UAV is rehearsed according to the UAV flight information in the flight model to obtain a rehearsal model.
7. The method for collaborative management of data information based on deep learning according to claim 1, characterized in that: The step of establishing corresponding information between the preview model and the real-time flight model, the task UAV terminal flying based on the UAV flight information, obtaining the trajectory of the task UAV terminal and transmitting it to the flight model in real time to obtain the real-time flight model includes: Obtaining the rehearsal time information corresponding to the UAV flight route information in the flight mission information in the rehearsal model, and matching the UAV flight route information in the rehearsal model with the identification points in the UAV flight route information of the flight model based on the rehearsal time information to obtain corresponding information; The task UAV is flown in the work area according to the UAV flight information and the real-time trajectory of the task UAV is obtained, and the real-time trajectory is constructed in the flight model to obtain a real-time flight model; The trajectory accuracy index is obtained based on the coincidence information between the real-time flight model and the preview model.
8. The method for collaborative management of data information based on deep learning according to claim 1, characterized in that: The step of using the real-time flight model corresponding to the trajectory accuracy index that does not meet the preset conditions as the impact model, and re-rehearsing the rehearsal model based on the impact model until all the mission drones complete the drone flight information, includes: Set preset conditions, and use the real-time flight model corresponding to the trajectory accuracy index that does not meet the preset conditions as the influencing model; Based on the impact model, the response rehearsal model is re-rehearsed to obtain an updated rehearsal model, and whether there is conflict information on the task UAV side in the updated rehearsal model, wherein the conflict information is a collision situation between the task UAV side and the UAV flight route information; If there is conflicting information on the mission drone side in the updated rehearsal model, the mission drone side will be adjusted in flight trajectory or flight speed until all mission drone sides complete the drone flight information.