Method, apparatus and transportation system for navigation based on a drone
By acquiring environmental data from drones and updating it in real time, and combining this with deep learning supervised learning methods to train a navigation model, the problem of insufficient navigation training data in AGV navigation is solved, thus improving the stability and accuracy of navigation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- YIWU QINGYUE OPTOELECTRONICS TECHNOLOGY INSTITUTE CO LTD
- Filing Date
- 2023-06-29
- Publication Date
- 2026-06-12
AI Technical Summary
Existing path planning technologies suffer from insufficient navigation training data in AGV navigation, leading to unstable navigation accuracy.
By acquiring environmental data through drones, estimating the time required for the task, matching action groups, and updating environmental data in real time to guide the AGV to complete the task, the navigation model is trained using deep learning supervised learning methods.
It improves the stability and accuracy of AGV navigation and makes up for the lack of navigation training data.
Smart Images
Figure CN116817922B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of automated guided vehicle control technology, and in particular to a method, apparatus and transportation system based on unmanned aerial vehicle (UAV) navigation. Background Technology
[0002] As an important automated handling robot in the production process, "Automated Guided Vehicle (AGV)" requires efficient path planning technology to support its automated transportation capabilities in order to improve automated production efficiency.
[0003] Currently, existing path planning technologies include traditional algorithms based on search optimization, machine learning algorithms, and deep learning algorithms. Among them, deep learning algorithms are further divided into different routes such as supervised learning and reinforcement learning.
[0004] However, due to the diverse navigation planning scenarios of AGVs on production lines and the limited navigation training data, the above-mentioned path planning technology suffers from insufficient navigation training data and unstable navigation accuracy. Summary of the Invention
[0005] This invention provides a method, apparatus, and transportation system for navigation based on unmanned aerial vehicles (UAVs), which solves the problems of insufficient navigation training data and unstable navigation accuracy in path planning technology, and can effectively improve the accuracy and stability of navigation.
[0006] According to one aspect of the present invention, a method for navigation based on an unmanned aerial vehicle (UAV) is provided, comprising:
[0007] The current environmental data is acquired by the drone; the current environmental data includes environmental obstacle map data corresponding to at least one preset time from the location of the automated guided vehicle to the destination, and the preset time includes at least one frame.
[0008] The estimated time required to execute the target task is based on the current environmental data; the estimated time includes at least two of the preset times.
[0009] Predict environmental data for the next preset time based on the current environmental data, so as to obtain environmental data for the required time.
[0010] The action group corresponding to the target task is determined based on the environmental data of the required time; the action group includes at least two individual actions.
[0011] Navigate according to the action group to complete the target task.
[0012] Optionally, predicting environmental data for the next preset time based on the current environmental data to obtain environmental data for the required time includes:
[0013] The current environmental data is matched with the preset environmental data for the next preset time, and used as the environmental data for the next preset time.
[0014] Iterate through all preset times corresponding to the required time to obtain the environmental data for the required time.
[0015] Optionally, the action group corresponding to the target task is determined based on the environmental data of the required time, including:
[0016] Match the environmental data for the required time and the corresponding preset action group according to the preset action group time axis;
[0017] Add the preset action group corresponding to the environmental data of the required time to the environmental data of the required time to form the action group corresponding to the target task.
[0018] Optionally, when completing the target task according to the action group navigation, the method further includes:
[0019] The drone acquires real-time environmental data of the current location of the automated guided vehicle.
[0020] The current action of the automated guided vehicle is corrected based on the environmental data of the current location of the automated guided vehicle.
[0021] Optionally, the current action of the automated guided vehicle (AGV) can be corrected based on environmental data of its current location, including:
[0022] Based on the environmental data of the current location of the automated guided vehicle, a preset single action attribute group is matched to determine the matched preset single action; the single action attribute group includes the action attribute of the single action and the environmental data of the corresponding time.
[0023] The current action is determined based on the matched preset single action and the single action corresponding to the action group, so as to correct the action group.
[0024] Optionally, the currently executed action is determined based on the matched preset single action and the single action corresponding to the action group, including:
[0025] Set the weights of the preset single action for matching and the single action corresponding to the action group;
[0026] The currently executing action is determined based on the matched preset single action and the single action corresponding to the action group, as well as their respective weights.
[0027] Optionally, before acquiring current environmental data via drone, the following may also be included:
[0028] The environmental data corresponding to the simulated path of the target mission of the automated guided vehicle is obtained by using a drone;
[0029] Based on the environmental data corresponding to the simulated path, determine the preset environmental data, preset action groups, and preset single action attribute groups;
[0030] Using the environmental data corresponding to the simulated path as training data, the preset environmental data, the preset action group, and the preset single action attribute group are trained based on a deep learning supervised learning method.
[0031] According to another aspect of the present invention, a device for navigation based on an unmanned aerial vehicle (UAV) is provided, comprising: a current data acquisition module, a time estimation module, a data acquisition module for the time required for task execution, an action group determination module, and a navigation module;
[0032] The current data acquisition module is used to acquire current environmental data through the drone; the current environmental data includes environmental obstacle map data corresponding to at least one preset time from the location of the automated guided vehicle to the destination, and the preset time includes at least one frame.
[0033] The time estimation module is used to estimate the time required to execute the target task based on the current environmental data; the required time includes at least two preset times.
[0034] The task execution time data acquisition module is used to predict the environmental data for the next preset time based on the current environmental data, so as to obtain the environmental data for the required time.
[0035] The action group determination module is used to determine the action group corresponding to the target task based on the environmental data of the required time; the action group includes at least two individual actions.
[0036] The navigation module is used to navigate according to the action group to complete the target task.
[0037] According to another aspect of the present invention, a transportation system is provided, including a drone and a drone-based navigation device as described in the embodiments of the present invention;
[0038] The drone and the device for navigation based on the drone are communicatively connected.
[0039] Optionally, the drone includes a 3D camera.
[0040] The technical solution provided by this invention involves pairing a drone with an AGV to acquire environmental data, matching the environmental data with the action set required for the AGV to complete the task, thereby achieving navigation of the target task and making up for the problem of insufficient navigation training data. By updating the environmental data in real time by the drone and using the updated environmental data as input to rematch the action set required for the AGV to complete the task, the stability of navigation is effectively improved.
[0041] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description
[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0043] Figure 1 This is a flowchart of a method for navigation based on an unmanned aerial vehicle (UAV) provided in Embodiment 1 of the present invention.
[0044] Figure 2 This is a flowchart of a method for navigation based on an unmanned aerial vehicle (UAV) according to Embodiment 2 of the present invention.
[0045] Figure 3 This is a flowchart of another method for navigation based on a drone, provided in Embodiment 2 of the present invention.
[0046] Figure 4 This is a flowchart of another method for navigation based on a drone, provided in Embodiment 2 of the present invention.
[0047] Figure 5 This is a schematic diagram of a device for navigation based on a drone, provided in Embodiment 3 of the present invention. Detailed Implementation
[0048] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0049] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0050] Example 1
[0051] Figure 1 This is a flowchart of a method for navigation based on a drone, provided in Embodiment 1 of the present invention. This embodiment is applicable to situations where path planning technology suffers from insufficient navigation training data and unstable navigation accuracy. The method can be executed by a device for navigation based on a drone, which can be implemented in hardware and / or software. This device can be configured in any electronic device with network communication capabilities. Figure 1 As shown, the method includes:
[0052] S110: Acquire current environmental data via drones.
[0053] Specifically, each AGV in the production workshop is assigned a corresponding drone. The AGV and the drone can share and transmit data through Robot Operating System (ROS), Transmission Control Protocol (TPC) or other communication methods. The drone and AGV share a charging node. The AGV is equipped with a controller for data reception, data storage, data processing, algorithm calculation, and sending control commands. The production workshop reserves a drone flight area at a designated height, higher than the height of personnel and mobile devices in the production process, to ensure unobstructed flight. The drone obtains the ground movement path corresponding to the AGV at its flight altitude through manual operation. During the flight, it collects images of ground equipment and personnel, and generates environmental obstacle map data for AGV operation by reconstructing a 3D production environment model containing depth information. The current environmental data includes environmental obstacle map data corresponding to at least one preset time from the location of the automated guided vehicle to the destination, with each preset time consisting of at least one frame. The environmental data consists of several frames of environmental obstacle map data, with each frame corresponding to a preset time. Specifically, the current environmental data can be understood as one frame of environmental obstacle map data corresponding to at least one preset time from the location of the automated guided vehicle to the destination. The preset time should be less than the time required to perform the target task, and the number of frames of environmental obstacle map data corresponding to each preset time is equal. The data acquisition operation of this UAV can be carried out simultaneously with production or in a specially designed experimental scenario, and this invention does not limit this.
[0054] S120. Estimate the time required to execute the target task based on current environmental data.
[0055] The required time includes at least two preset times. The target task can be specifically understood as the process of an automated guided vehicle moving an object to its destination according to the needs of the staff.
[0056] Specifically, the time required to execute the target task is divided into segments into several time periods. Each time period corresponds to one or more frames of environmental obstacle maps. Based on the time period corresponding to the current environmental data, the environmental obstacle map corresponding to the preset time is obtained. The environmental obstacle map corresponding to the preset time is used as the input for the new environmental obstacle map. Then, the environmental obstacle map corresponding to the next preset time is obtained. The environmental obstacle map corresponding to the next preset time is matched with the current time period to determine the current environmental obstacle map. This method is used until the time required to execute the target task is traversed.
[0057] S130. Predict the environmental data for the next preset time based on the current environmental data, so as to obtain the environmental data for the required time.
[0058] Specifically, the time required to execute the target task is divided into several time periods, each corresponding to one or more frames of environmental obstacle maps. Based on the time period corresponding to the current environmental data, the environmental obstacle map corresponding to a preset time is obtained. The environmental obstacle map corresponding to the preset time is used as the input for the new environmental obstacle map. Then, the environmental obstacle map corresponding to the next preset time is obtained. The environmental obstacle map corresponding to the next preset time is matched with the current time period to determine the current environmental obstacle map. This method is used until all environmental obstacle maps within the time range required to execute the target task are traversed to obtain the environmental data for the time required to execute the target task.
[0059] S140. Determine the action group corresponding to the target task based on the environmental data of the required time.
[0060] Specifically, a complete AGV transport task performed by a drone includes a set of motion vectors to complete the task. Using environmental data representing the time required for task execution as input, the motion vectors are matched against the environmental data for the required time and the preset motion sets corresponding to each time segment within the time axis. The preset motion sets corresponding to each time segment are then added to the environmental data for the required time to determine the motion set corresponding to the target task. Each motion set includes at least two individual actions, each labeled with parameters such as acceleration, angular acceleration, angular velocity, angle, speed, state duration, and time axis.
[0061] For example, the action vector group can be [A(..); F(..); D(..); S(..); T(..)], where A represents acceleration, F represents maintaining straight line, D represents deceleration, S represents stopping, and T represents turning.
[0062] S150. Complete the target task according to the action group navigation.
[0063] Specifically, the drone acquires current environmental data, estimates the time required to execute the target task based on the current environmental data, and predicts the environmental data for the next preset time based on the current environmental data, so as to obtain the environmental data for the required time. Based on the environmental data for the required time, the action group corresponding to the target task is determined to realize path planning, and the automated guided vehicle is navigated to complete the execution of the target task.
[0064] The technical solution provided by this invention involves pairing a drone with an AGV to acquire environmental data, matching the environmental data with the action set required for the AGV to complete the task, thereby achieving navigation of the target task and making up for the problem of insufficient navigation training data. By updating the environmental data in real time by the drone and using the updated environmental data as input to rematch the action set required for the AGV to complete the task, the stability of navigation is effectively improved.
[0065] Example 2
[0066] Figure 2 This is a flowchart illustrating a method for navigation based on an unmanned aerial vehicle (UAV) according to Embodiment 2 of the present invention. This embodiment further refines the aforementioned embodiments. For example... Figure 2 As shown, the method includes:
[0067] S210: Acquire current environmental data via drones.
[0068] S220. Estimate the time required to execute the target task based on current environmental data.
[0069] S230. Match the preset environmental data for the next preset time with the current environmental data, and use it as the environmental data for the next preset time.
[0070] The preset environmental data includes environmental obstacle map data for a complete simulation of the drone carrying out the AGV transport task. Specifically, the time required to perform the target task is divided into several time periods, each time period corresponding to one or more frames of environmental obstacle map. Based on the time period corresponding to the current environmental data, the environmental obstacle map corresponding to the preset time is obtained, and the environmental obstacle map corresponding to the preset time is used as the environmental obstacle map for the next preset time.
[0071] S240. Iterate through all preset times corresponding to the required time and obtain the environmental data for the required time.
[0072] Specifically, the environmental obstacle map for the next preset time is taken as input. Based on the environmental obstacle map for the next preset time, the environmental obstacle map corresponding to the preset time after the next preset time is obtained. The environmental obstacle map corresponding to the preset time is matched with the environmental obstacle maps corresponding to the preset times within the time range required to execute the target task. This method is used until all environmental obstacle maps within the time range required to execute the target task are traversed to obtain all environmental data required for the execution of the target task.
[0073] S250: Match the required time environment data and the corresponding time preset action group according to the preset action group time axis.
[0074] The preset action group includes the actions performed by the drone in a complete simulation of an AGV transport task, including turning, stopping, accelerating, decelerating, and maintaining straight-line movement. Each action includes parameter annotations such as acceleration, angular acceleration, angular velocity, angle, speed, state duration, and time axis.
[0075] Specifically, a complete simulation of an AGV transport task by a drone includes a set of motion vectors to complete the task. The environmental data required for the task execution is used as input, and the environmental data required for the time is matched with the preset motion sets corresponding to each time segment within the time range through the motion vector set time axis.
[0076] S260. Add the preset action group corresponding to the environmental data of the required time to the environmental data of the required time to form the action group corresponding to the target task.
[0077] Specifically, preset action groups corresponding to each time segment within the time axis range required for the execution of the target task are added to the environmental data for the required time, thereby determining the action groups corresponding to the target task.
[0078] S270: When completing the target task according to the action group navigation, the UAV acquires the environmental data of the current position of the automated guided transport vehicle in real time.
[0079] Specifically, during the execution of a task action group by the automated guided vehicle, the drone acquires a real-time map of the changing environmental obstacles and uses this map as input as environmental data for the current location of the automated guided vehicle.
[0080] S280. Correct the current action of the automated guided vehicle based on the environmental data of the current location of the automated guided vehicle.
[0081] Specifically, the environmental data of the current location of the automated guided vehicle is used as input. Based on the environmental data of the current location of the automated guided vehicle, a preset single action attribute group is matched to determine the matched preset single action. Based on the matched preset single action and the single action corresponding to the action group, the current action to be executed is determined, so as to correct the action group and complete the navigation task.
[0082] Figure 3 A flowchart of another method for navigation based on a drone provided in Embodiment 2 of the present invention is shown below. Figure 3 As shown, the method includes:
[0083] S310: Acquire current environmental data via drones.
[0084] S320. Estimate the time required to execute the target task based on current environmental data.
[0085] S330. Match the preset environmental data for the next preset time with the current environmental data, and use it as the environmental data for the next preset time.
[0086] S340. Iterate through all preset times corresponding to the required time and obtain the environmental data for the required time.
[0087] S250: Match the required time environment data and the corresponding time preset action group according to the preset action group time axis.
[0088] S360: Add the preset action group corresponding to the environmental data of the required time to the environmental data of the required time to form the action group corresponding to the target task.
[0089] S370: When completing the target task according to the action group navigation, the UAV acquires the environmental data of the current position of the automated guided transport vehicle in real time.
[0090] S380. Match a preset single action attribute group with the environmental data of the current position of the automated guided vehicle, and determine the matched preset single action.
[0091] Specifically, the system matches the motion attributes of a preset single action with the environmental data of the current location of the automated guided vehicle (AGV) and the environmental data of the corresponding time for that single action. The matching preset single action is determined by whether the motion attributes of the preset single action and the environmental data of the corresponding time for that single action are consistent. The single action attribute group includes the motion attributes of the single action and the environmental data of the corresponding time; the motion attributes include acceleration, angular acceleration, angular velocity, angle, velocity, state duration, and time axis.
[0092] S390. Determine the currently executed action based on the matched preset single action and the single action corresponding to the action group, so as to correct the action group.
[0093] Specifically, action weights are assigned to the matching preset single actions and action groups, and the current action to be executed is determined by the matching preset single actions and action groups and their respective weights. The task action groups that were originally to be executed are then corrected to complete the navigation task.
[0094] Optionally, the current action to be executed is determined based on the matched preset single action and the single action corresponding to the action group, including:
[0095] Set the weights of the preset individual actions and action groups for matching;
[0096] The current action is determined based on the matched preset single action and the single action corresponding to the action group, as well as their respective weights.
[0097] The technical solution provided by this invention decomposes the action group required for the task into individual actions, and matches the individual actions with environmental data, which effectively improves the accuracy of navigation. Based on the matched preset individual actions and the individual actions corresponding to the action group, the current action to be executed is determined to correct the action group, thereby further improving the accuracy of navigation.
[0098] Figure 4 A flowchart of another method for navigation based on an unmanned aerial vehicle (UAV) provided in Embodiment 2 of the present invention is shown below. Figure 4 As shown, optionally, before acquiring current environmental data via drone, the following steps are also included:
[0099] S410: Obtain environmental data corresponding to the simulated path of the target mission of the automated guided vehicle through the drone.
[0100] Specifically, by simulating the movement trajectory of AGV (Automated Guided Vehicle) transport robots using drones, navigation model training data can be collected in various scenarios, including actual production situations. By selecting drones equipped with 3D cameras to simulate the movement trajectory of automated guided vehicles (AGVs), depth information of environmental obstacles can be collected, forming a 3D reconstruction of the obstacles and generating an environmental obstacle map. The specific method involves: the drone being manually operated to simulate the ground movement path of the AGV at its flight altitude, collecting images of ground equipment and personnel during flight, and constructing a production environment model containing depth information through 3D reconstruction, thus creating a real-time map of obstacles in the AGV's operating environment. This drone data acquisition operation can be performed simultaneously with production or in specially designed experimental scenarios; this invention does not limit the scope of the data acquisition.
[0101] S420. Determine the preset environment data, preset action group, and preset single action attribute group based on the environment data corresponding to the simulation path.
[0102] Specifically, the movement path of the simulated automated guided vehicle (AGV) is labeled using different methods. The environmental obstacle map within the time range required for the target task is segmented into groups of several frames. By constructing paired labeling data between the current environmental obstacle map group and subsequent environmental obstacle map groups, the map data of the previous segment and the next segment are matched into a set of labeled data, which serves as the preset environmental data. The data of a complete simulation of an AGV transport task is labeled, including the motion vector group for completing the task. Within the range corresponding to the time axis attribute of the motion group, a motion environmental obstacle map corresponding to the task changing over time is added as paired data. This is done by constructing the complete motion vector group required for the AGV to complete the task. The labeled data formed by pairing the action group with the corresponding environmental obstacle map frame group serves as a preset action group, and the data format is a multi-frame environmental obstacle map. The action group required to execute the target task is decomposed, and each action includes a time range from the end of the previous action to the completion of the current action. The environmental obstacle map that changes over time within this time range is matched as the pairing data, and the attributes of the action itself are merged into a single action attribute group. Whether the task is terminated (collision, obstacle, etc.) after the end of a single action and before the start of the next action is used as feedback on whether the task is successful or not. That is, when a collision or obstacle occurs, a task termination feedback signal will be displayed, and when no collision or obstacle occurs, a task success feedback signal will be displayed. This completes the labeling of one action data and forms a preset single action attribute group.
[0103] S430. Using the environmental data corresponding to the simulated path as training data, train the preset environmental data, preset action groups, and preset single action attribute groups based on the deep learning supervised learning method.
[0104] Specifically, different labeled data collected by drone simulation are used as training data and trained using deep learning supervised learning methods. First, the network is trained on preset environmental data, with the goal of predicting map data for the next few frames from the previous few frames. The network contains fully convolutional layers and outputs several frames of map data, resulting in the first training model. Second, the network is trained on preset action groups. The network contains fully connected layers and outputs action groups, each action containing attributes in the form of a matrix, resulting in the second training model. Finally, the network is trained on preset single action attribute groups. The network contains fully connected layers and outputs actions, each action containing attributes in the form of a vector, resulting in the third training model.
[0105] S440: Acquires current environmental data via drones.
[0106] S450. Estimate the time required to execute the target task based on current environmental data.
[0107] S460. Predict environmental data for the next preset time based on current environmental data, so as to obtain environmental data for the required time.
[0108] S470. Determine the action group corresponding to the target task based on the environmental data of the required time.
[0109] S480. Complete the target task according to the action group navigation.
[0110] The technical solution provided in this invention collects navigation training data by simulating the movement trajectory of an automated guided vehicle (AGV) using a drone. This data can be used to decompose the AGV's movements and paired with collected environmental obstacle maps to form a standard dataset. By training a model, a supervised learning prediction model is used to predict and guide the AGV's path planning, enabling it to complete the target task. This overcomes the problem of insufficient navigation training data and effectively improves navigation stability and accuracy.
[0111] Example 3
[0112] Figure 5 This is a schematic diagram of a device for navigation based on a drone, as provided in Embodiment 3 of the present invention. Figure 5 As shown, the device includes: a current data acquisition module 510, a time estimation module 520, a data acquisition module 530 for the time required for task execution, an action group determination module 540, and a navigation module 550;
[0113] The current data acquisition module 510 is used to acquire current environmental data through the drone; the current environmental data includes environmental obstacle map data corresponding to at least one preset time from the location of the automated guided vehicle to the destination, and the preset time includes at least one frame;
[0114] The time estimation module 520 is used to estimate the time required to execute the target task based on the current environmental data; the required time includes at least two preset times;
[0115] The task execution time data acquisition module 530 is used to predict the environmental data for the next preset time based on the current environmental data, so as to obtain the environmental data for the required time.
[0116] The action group determination module 540 is used to determine the action group corresponding to the target task based on the environmental data of the required time; the action group includes at least two individual actions.
[0117] The navigation module 550 is used to navigate according to the action group to complete the target task.
[0118] The device for navigation based on unmanned aerial vehicles (UAVs) provided in the embodiments of the present invention can execute the navigation method based on UAVs provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the method.
[0119] This invention also provides a transportation system, including a drone and a device for navigation based on the drone provided in this invention; wherein the drone and the device for navigation based on the drone are communicatively connected; the drone includes a three-dimensional camera.
[0120] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.
[0121] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.
Claims
1. A method for navigation based on unmanned aerial vehicles (UAVs), characterized in that, include: The current environmental data is acquired by the drone; the current environmental data includes environmental obstacle map data corresponding to at least one preset time from the location of the automated guided vehicle to the destination, and the preset time includes at least one image frame. The estimated time required to execute the target task is based on the current environmental data; the estimated time includes at least two of the preset times. Predict environmental data for the next preset time based on the current environmental data, so as to obtain environmental data for the required time. The action group corresponding to the target task is determined based on the environmental data of the required time; the action group includes at least two individual actions. Complete the target task according to the action group navigation; Based on the environmental data of the required time, determine the action group corresponding to the target task, including: The environmental data for the required time and the corresponding preset action group are matched according to the preset action group time axis; wherein, the preset action group is the action performed by the drone to completely simulate an AGV handling task, including turning, stopping, accelerating, decelerating and maintaining straight movement. Each action has action attributes, which include acceleration, angular acceleration, angular velocity, angle, speed, state duration, and time axis. Add the preset action group corresponding to the environmental data of the required time to the environmental data of the required time to form the action group corresponding to the target task; When completing the target task according to the action group navigation, it also includes: The drone acquires real-time environmental data of the current location of the automated guided vehicle. The current action of the automated guided vehicle is corrected based on the environmental data of the current location of the automated guided vehicle; The current action of the automated guided vehicle (AGV) is corrected based on environmental data of its current location, including: Based on the environmental data of the current location of the automated guided vehicle, a preset single action attribute group is matched to determine the matched preset single action; the single action attribute group includes the action attribute of the single action and the environmental data of the corresponding time. The current action to be executed is determined based on the matched preset single action and the single action corresponding to the action group, so as to correct the action group; The current action to be executed is determined based on the matched preset single action and the single action corresponding to the action group, including: Set the weights of the preset single action for matching and the single action corresponding to the action group; The currently executing action is determined based on the matched preset single action and the single action corresponding to the action group, as well as their respective weights.
2. The method according to claim 1, characterized in that, Based on the current environmental data, predict the environmental data for the next preset time to obtain the environmental data for the required time, including: The current environmental data is matched with the preset environmental data for the next preset time, and used as the environmental data for the next preset time. Iterate through all preset times corresponding to the required time to obtain the environmental data for the required time.
3. The method according to claim 1, characterized in that, Before acquiring current environmental data via drones, the following is also included: The environmental data corresponding to the simulated path of the target mission of the automated guided vehicle is obtained by using a drone; Based on the environmental data corresponding to the simulated path, determine the preset environmental data, preset action groups, and preset single action attribute groups; Using the environmental data corresponding to the simulated path as training data, the preset environmental data, the preset action group, and the preset single action attribute group are trained based on a supervised learning method.
4. A device for navigation based on unmanned aerial vehicles (UAVs), characterized in that, The apparatus for performing a method for navigation based on a drone as described in any one of claims 1-3 includes: a current data acquisition module, a time estimation module, a data acquisition module for the time required for task execution, an action group determination module, and a navigation module; The current data acquisition module is used to acquire current environmental data through the drone; the current environmental data includes environmental obstacle map data corresponding to at least one preset time from the location of the automated guided vehicle to the destination, and the preset time includes at least one frame. The time estimation module is used to estimate the time required to execute the target task based on the current environmental data; the required time includes at least two preset times. The task execution time data acquisition module is used to predict the environmental data for the next preset time based on the current environmental data, so as to obtain the environmental data for the required time. The action group determination module is used to determine the action group corresponding to the target task based on the environmental data of the required time; the action group includes at least two individual actions. The navigation module is used to navigate according to the action group to complete the target task.
5. A transportation system, characterized in that, Includes drones and the drone-based navigation device as described in claim 4; The drone and the device for navigation based on the drone are communicatively connected.
6. The transportation system according to claim 5, characterized in that, The drone includes a 3D camera.