A method, apparatus, equipment and storage medium for unmanned aerial vehicle (UAV) flight path planning

Abstract

This application discloses a method, apparatus, device, and storage medium for unmanned aerial vehicle (UAV) route planning. It obtains the target UAV's model, takeoff location information, and destination information from the target UAV's takeoff request information. Based on the takeoff location and destination information, it determines a set of routes within the airspace altitude. From this set, it obtains the occupancy rate, distance, and weather information for each route. Based on the occupancy rate and distance, it determines candidate routes for the target UAV. These candidate routes are then input into a route selection model to obtain the target route. Compared to existing technologies, this application first filters the route set using takeoff location and destination information, then determines candidate routes based on occupancy rate and weather information, and inputs these candidate routes into the route selection model to obtain the target route. By using real-time weather information and the route selection model, it determines the optimal target route, thereby improving route safety while ensuring flight efficiency.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicles (UAVs), and in particular to a method, apparatus, equipment, and storage medium for UAV route planning. Background Technology

[0002] Drone technology has made significant progress in recent years, and drones have become a widely used robotic technology. Flight path planning plays a crucial role in ensuring the safe and efficient completion of drone flight missions, determining both flight efficiency and safety.

[0003] Currently, drone flight path design mainly relies on pre-setting by human subjective experience and GPS navigation. Pre-setting by human subjective experience involves determining which areas a drone can fly in based on extensive practical experience flying drones, and then setting the flight path in the drone's program. The disadvantages of this method are that it requires a lot of time and manpower, is costly, and is not suitable for large-scale deployment in open environments.

[0004] GPS navigation determines the drone's location and calculates its flight path using satellite positioning. This method can be used in open environments and is relatively inexpensive. However, GPS signals may be subject to interference, and the flight altitude and potential weather conditions cannot be determined, resulting in lower flight safety.

[0005] To solve these problems, it is necessary to invent a new method for drone flight path planning. Summary of the Invention

[0006] In view of this, this application provides a method, apparatus, device and storage medium for unmanned aerial vehicle (UAV) flight path planning, which can improve flight efficiency and safety.

[0007] To achieve the above objectives, the following solution is proposed:

[0008] A method for planning unmanned aerial vehicle (UAV) flight paths, comprising:

[0009] Obtain takeoff request information from the target drone, the takeoff request information including the model of the target drone, takeoff location information and destination information;

[0010] Determine the corresponding airspace altitude based on the model of the target UAV;

[0011] Based on the departure location information and destination information, determine the set of flight routes in the airspace altitude;

[0012] Obtain the occupancy rate, route distance, and weather information for each route in the route set;

[0013] Based on the occupancy rate and distance of each route, the candidate routes for the target UAV are determined;

[0014] Input the candidate routes and corresponding meteorological information into the route selection model to obtain the target route output by the route selection model.

[0015] The route selection model is trained using historical routes and corresponding historical meteorological information as training samples, and target routes in the historical routes as training labels.

[0016] Optionally, the route selection model includes an input layer, a convolutional layer, a pooling layer, a prediction layer, and an output layer;

[0017] The training process of the route selection model includes:

[0018] Historical flight routes and corresponding historical weather information are obtained through the input layer;

[0019] By using convolutional layers, features are extracted from the historical flight routes and corresponding historical meteorological information to obtain a feature matrix;

[0020] The feature matrix is ​​pooled using a pooling layer to obtain a feature matrix of uniform size.

[0021] The prediction layer uses a feature matrix of uniform size to predict the weather for each historical flight path, thus obtaining the predicted weather.

[0022] Based on the predicted weather, the target route in the historical flight path is determined through the output layer.

[0023] The route loss value is determined based on the difference between the target route result in the determined historical routes and the target route label in the historical routes.

[0024] The parameters of the route selection model are updated based on the route loss value.

[0025] Optionally, determining the corresponding airspace altitude based on the model of the target UAV includes:

[0026] The corresponding airspace altitude is determined based on the pre-established airspace division rules and the model of the target UAV.

[0027] Optionally, based on the departure location information and destination information, a set of flight routes in the airspace altitude is determined, including:

[0028] Based on the departure and destination information, determine the battery replenishment stations along each route;

[0029] The routes with fewer than a preset threshold of battery replenishment stations are grouped into a set of routes in the airspace altitude.

[0030] A drone flight path planning device, comprising:

[0031] The first information acquisition module is used to acquire takeoff request information of the target UAV, the takeoff request information including the model of the target UAV, takeoff location information and destination information;

[0032] An altitude determination module is used to determine the corresponding airspace altitude based on the model of the target UAV;

[0033] The set determination module is used to determine the set of flight routes in the airspace altitude based on the takeoff location information and destination information;

[0034] The second information acquisition module is used to acquire the occupancy rate, route distance, and weather information of each route in the route set;

[0035] The candidate route determination module is used to determine the candidate routes for the target UAV based on the occupancy rate and route distance of each route.

[0036] The target route output module is used to input the candidate route and corresponding meteorological information into the route selection model to obtain the target route output by the route selection model.

[0037] The route selection model is trained using historical routes and corresponding historical meteorological information as training samples, and target routes in the historical routes as training labels.

[0038] Optionally, the device further includes a model training module, which includes an input layer, a convolutional layer, a pooling layer, a prediction layer, and an output layer.

[0039] The model training module is used to obtain historical flight routes and corresponding historical meteorological information through the input layer;

[0040] By using convolutional layers, features are extracted from the historical flight routes and corresponding historical meteorological information to obtain a feature matrix;

[0041] The feature matrix is ​​pooled using a pooling layer to obtain a feature matrix of uniform size.

[0042] The prediction layer uses a feature matrix of uniform size to predict the weather for each historical flight path, thus obtaining the predicted weather.

[0043] Based on the predicted weather, the target route in the historical flight path is determined through the output layer.

[0044] The route loss value is determined based on the difference between the target route result in the determined historical routes and the target route label in the historical routes.

[0045] The parameters of the route selection model are updated based on the route loss value.

[0046] Optionally, the height determination module includes:

[0047] The altitude determination submodule is used to determine the corresponding airspace altitude based on the pre-defined airspace division rules and the model of the target UAV.

[0048] Optionally, the set determination module includes:

[0049] The first set determination submodule is used to determine battery replenishment stations along each route based on the departure location information and destination information.

[0050] The second set determination submodule is used to form a set of routes in the airspace altitude where the number of battery replenishment stations is less than a preset threshold.

[0051] A drone flight path planning device, comprising a memory and a processor;

[0052] The memory is used to store programs;

[0053] The processor is used to execute the program to implement each step of the aforementioned UAV route planning method.

[0054] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the various steps of the aforementioned UAV route planning method.

[0055] As can be seen from the above technical solutions, the UAV route planning method, apparatus, equipment, and storage medium provided in this application obtain the target UAV model, takeoff location information, and destination information from the target UAV takeoff request information. Based on the takeoff location information and destination information, a set of routes in the airspace altitude is determined. Then, from the route set, the occupancy rate, route distance, and meteorological information of each route are obtained. Based on the occupancy rate and route distance of each route, candidate routes for the target UAV are further determined. Finally, the candidate routes are input into the route selection model to obtain the target route. Compared with the prior art, this application first filters out the route set through takeoff location information and destination information, then determines candidate routes from the route set based on occupancy rate and meteorological information, and inputs the candidate routes into the route selection model to obtain the target route. By obtaining real-time meteorological information and the route selection model, the optimal target route is determined, which improves route safety while ensuring flight efficiency. Attached Figure Description

[0056] Figure 1 A flowchart of a UAV route planning method provided in this application embodiment;

[0057] Figure 2A flowchart of a route selection model training method provided in this application embodiment;

[0058] Figure 3 This is a schematic diagram of the structure of a drone route planning device provided in an embodiment of this application;

[0059] Figure 4 This is a hardware structure block diagram of a drone route planning device disclosed in an embodiment of this application. Detailed Implementation

[0060] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0061] Figure 1 A method for planning unmanned aerial vehicle (UAV) routes, provided in this application embodiment, may include the following steps:

[0062] Step S100: Obtain the takeoff request information of the target UAV.

[0063] Specifically, the takeoff request information may include the target drone's model, takeoff location information, and destination information. This information can be obtained through the drone control platform or by connecting to the internet, and can be obtained through computers, mobile phones, or other electronic devices with similar functions.

[0064] Step S110: Determine the corresponding airspace altitude based on the model of the target UAV.

[0065] Specifically, to ensure smooth air traffic and improve the efficiency of air route utilization, different drone models can have different flight altitudes, and the corresponding airspace altitude can be determined according to the drone model.

[0066] Step S120: Based on the takeoff location information and destination information, determine the set of flight routes in the airspace altitude.

[0067] Specifically, there can be multiple routes from the takeoff point to the destination. Different routes may have the same destination, and the airspace altitudes traversed by each route may be different. The airspace altitudes in which the same UAV flies may be the same. Based on the takeoff point and destination information of the UAV, the routes in the specified airspace altitudes can be determined and formed into a route set.

[0068] Step S130: Obtain the occupancy rate, route distance, and weather information of each route in the route set.

[0069] Specifically, the occupancy rate, route distance, and weather information of each route can be different. The occupancy rate of a route can represent the number of drones occupying that route at the current moment, the route distance can represent the length of the route traversed by each route, and the weather information can represent the weather conditions in the airspace that the route traverses at the current moment. Based on the set of routes, the occupancy rate, route distance, and weather information of each corresponding route can be obtained.

[0070] Step S140: Determine the candidate routes for the target UAV based on the occupancy rate and route distance of each route.

[0071] Specifically, based on the occupancy rate and distance of each route, routes with lower occupancy rates and shorter distances can be selected as candidate routes. There can be one or more candidate routes.

[0072] Step S150: Input the candidate route and corresponding meteorological information into the route selection model to obtain the target route output by the route selection model.

[0073] Specifically, the route selection model is trained using historical routes and corresponding historical meteorological information as training samples, and target routes from the historical routes as training labels. Historical meteorological information can determine the meteorological stability of the airspace. Based on historical meteorological information, airspaces with relatively stable meteorological changes and corresponding historical routes within the airspace can be identified. In other words, the route selection model can select routes with more stable meteorological changes from the candidate routes as target routes to improve route safety. The route selection model can be a neural network model, and the structure of the neural network model can be varied, such as Convolutional Neural Network (CNN), BackPropagation Network (BP), RecurrentAttentionConvolutionalNeural Network (RA-CNN), and CenterNet, or an improved network structure of an existing network. The candidate routes and corresponding meteorological information can be input into the trained route selection model to obtain the target route output by the route selection model.

[0074] The above embodiment first filters out a set of routes based on takeoff and destination information. Then, from the set of routes, routes with low occupancy and good weather are selected as candidate routes based on occupancy rate and meteorological information. The candidate routes are input into the route selection model to obtain the target route. By using the acquired real-time meteorological information and the route selection model to determine the optimal target route, flight efficiency can be ensured, that is, the flight mission is completed within the specified flight time, while improving flight safety.

[0075] In some embodiments of this application, the route selection model in step S150 is described below. The route selection model may include an input layer, a convolutional layer, a pooling layer, a prediction layer, and an output layer, such as... Figure 2 As shown, the training process of the route selection model may include the following steps:

[0076] Step S151: Obtain historical flight routes and corresponding historical meteorological information through the input layer.

[0077] Step S152: Through convolutional layers, feature extraction is performed on historical flight routes and corresponding historical meteorological information to obtain a feature matrix.

[0078] Step S153: The feature matrix is ​​pooled through a pooling layer to obtain a feature matrix of uniform size.

[0079] Step S154: Through the prediction layer, based on a feature matrix of uniform size, the weather of each historical flight route is predicted to obtain the predicted weather.

[0080] Step S155: Through the output layer, determine the target route from the historical routes based on the predicted weather.

[0081] Step S156: Determine the route loss value based on the difference between the target route result in the determined historical routes and the target route label in the historical routes.

[0082] Step S157: Update the parameters of the route selection model based on the route loss value.

[0083] In the above embodiments, this application can predict the future weather conditions of the input route through a route selection model. By predicting and selecting routes in stable weather areas, the safety of the route can be improved.

[0084] In some embodiments of this application, the process of step S110, determining the corresponding airspace altitude based on the model of the target UAV, is described below. This process may include:

[0085] Step S111: Based on the departure location information and destination information, determine the battery replenishment stations along each route.

[0086] Specifically, the route points and battery replenishment stations for each route can be different. The corresponding routes and battery replenishment stations for each route can be determined based on the departure and destination information.

[0087] Step S112: Form a set of routes in the airspace altitude where the number of battery replenishment stations is less than a preset threshold.

[0088] Specifically, the more battery replenishment stations a route has, the longer the route. From all routes, a suitable length can be selected based on the number of battery replenishment stations. The preset threshold can be set based on the distance from each flyby to the destination. Among the routes that meet the airspace altitude requirements, routes with fewer than the preset threshold are further grouped into a route set. For example, if you take off from point A and go to point B, and the preset threshold is 10, routes with 9, 10, and 11 battery replenishment stations can be selected, and all routes meet the airspace altitude requirements. Based on the preset threshold, routes with 9 or 10 battery replenishment stations can be selected to form a route set.

[0089] In some embodiments of this application, an application scenario of a UAV route planning method is described below:

[0090] The first layer initially filters the first set of routes based on takeoff and destination information. The second set of routes is then filtered based on airspace altitude. The third set of routes is filtered based on the number of battery replenishment stations. The fourth set of routes is then filtered based on the occupancy rate, route distance, and weather information of each route in the third set. The fourth set of routes and the corresponding weather information are then input into the route selection model to obtain the target route. This application obtains the target route through a series of filters, which can improve flight efficiency and safety.

[0091] The UAV route planning device provided in the embodiments of this application is described below. The UAV route planning device described below and the UAV route planning method described above can be referred to each other.

[0092] like Figure 3 As shown, a schematic diagram of a drone route planning device is disclosed, which may include:

[0093] The first information acquisition module 11 is used to acquire takeoff request information of the target UAV, the takeoff request information including the model of the target UAV, takeoff location information and destination information;

[0094] Altitude determination module 12 is used to determine the corresponding airspace altitude based on the model of the target UAV;

[0095] The set determination module 13 is used to determine the set of flight routes in the airspace altitude based on the takeoff location information and destination information;

[0096] The second information acquisition module 14 is used to acquire the occupancy rate, route distance and meteorological information of each route in the route set;

[0097] The candidate route determination module 15 is used to determine the candidate routes of the target UAV based on the occupancy rate and route distance of each route.

[0098] The target route output module 16 is used to input the candidate route and corresponding meteorological information into the route selection model to obtain the target route output by the route selection model.

[0099] The route selection model is trained using historical routes and corresponding historical meteorological information as training samples, and target routes in the historical routes as training labels.

[0100] Optionally, the device further includes a model training module, which includes an input layer, a convolutional layer, a pooling layer, a prediction layer, and an output layer.

[0101] The model training module is used to obtain historical flight routes and corresponding historical meteorological information through the input layer;

[0102] By using convolutional layers, features are extracted from the historical flight routes and corresponding historical meteorological information to obtain a feature matrix;

[0103] The feature matrix is ​​pooled using a pooling layer to obtain a feature matrix of uniform size.

[0104] The prediction layer uses a feature matrix of uniform size to predict the weather for each historical flight path, thus obtaining the predicted weather.

[0105] Based on the predicted weather, the target route in the historical flight path is determined through the output layer.

[0106] The route loss value is determined based on the difference between the target route result in the determined historical routes and the target route label in the historical routes.

[0107] The parameters of the route selection model are updated based on the route loss value.

[0108] Optionally, the height determination module 12 includes:

[0109] The altitude determination submodule is used to determine the corresponding airspace altitude based on the pre-defined airspace division rules and the model of the target UAV.

[0110] Optionally, the set determination module 13 includes:

[0111] The first set determination submodule is used to determine battery replenishment stations along each route based on the departure location information and destination information.

[0112] The second set determination submodule is used to form a set of routes in the airspace altitude where the number of battery replenishment stations is less than a preset threshold.

[0113] The UAV route planning device provided in this application embodiment can be applied to UAV route planning equipment. The UAV route planning equipment can be a terminal. Figure 4 The hardware structure block diagram of the UAV route planning equipment is shown below. Figure 4 The hardware structure of the UAV route planning equipment may include: at least one processor 1, at least one communication interface 2, at least one memory 3, and at least one communication bus 4;

[0114] In this embodiment of the application, the number of processor 1, communication interface 2, memory 3, and communication bus 4 is at least one, and processor 1, communication interface 2, and memory 3 communicate with each other through communication bus 4;

[0115] Processor 1 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention.

[0116] Memory 3 may include high-speed RAM, and may also include non-volatile memory, such as at least one disk storage device;

[0117] The memory stores a program, which the processor can call. The program is used for:

[0118] Obtain takeoff request information from the target drone, the takeoff request information including the model of the target drone, takeoff location information and destination information;

[0119] Determine the corresponding airspace altitude based on the model of the target UAV;

[0120] Based on the departure location information and destination information, determine the set of flight routes in the airspace altitude;

[0121] Obtain the occupancy rate, route distance, and weather information for each route in the route set;

[0122] Based on the occupancy rate and distance of each route, the candidate routes for the target UAV are determined;

[0123] Input the candidate routes and corresponding meteorological information into the route selection model to obtain the target route output by the route selection model.

[0124] The route selection model is trained using historical routes and corresponding historical meteorological information as training samples, and target routes in the historical routes as training labels.

[0125] Optionally, the refined and extended functions of the program can be found in the description above.

[0126] This application embodiment also provides a storage medium that can store a program suitable for execution by a processor, the program being used for:

[0127] Obtain takeoff request information from the target drone, the takeoff request information including the model of the target drone, takeoff location information and destination information;

[0128] Determine the corresponding airspace altitude based on the model of the target UAV;

[0129] Based on the departure location information and destination information, determine the set of flight routes in the airspace altitude;

[0130] Obtain the occupancy rate, route distance, and weather information for each route in the route set;

[0131] Based on the occupancy rate and distance of each route, the candidate routes for the target UAV are determined;

[0132] Input the candidate routes and corresponding meteorological information into the route selection model to obtain the target route output by the route selection model.

[0133] The route selection model is trained using historical routes and corresponding historical meteorological information as training samples, and target routes in the historical routes as training labels.

[0134] Optionally, the refined and extended functions of the program can be found in the description above.

[0135] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0136] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined with each other, and the same or similar parts can be referred to each other.

[0137] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for planning flight routes for unmanned aerial vehicles (UAVs), characterized in that, The method comprises the following steps: obtaining take-off request information of a target unmanned aerial vehicle, the take-off request information comprising a model of the target unmanned aerial vehicle, take-off location information and destination information; determining a corresponding airspace height according to the model of the target unmanned aerial vehicle; determining a set of air routes in the airspace height based on the take-off location information and the destination information; obtaining an occupancy rate, a route distance and weather information of each air route in the set of air routes; determining a candidate air route of the target unmanned aerial vehicle according to the occupancy rate and the route distance of each air route; inputting the candidate air route and corresponding weather information into an air route selection model to obtain a target air route output by the air route selection model; the air route selection model is trained by taking historical air routes and corresponding historical weather information as training samples and taking a target air route in the historical air routes as a training label; wherein the air route selection model comprises an input layer, a convolution layer, a pooling layer, a prediction layer and an output layer; the training process of the air route selection model comprises: obtaining historical air routes and corresponding historical weather information through the input layer; extracting features of the historical air routes and corresponding historical weather information through the convolution layer to obtain a feature matrix; pooling the feature matrix through the pooling layer to obtain a feature matrix of uniform size; predicting weather of each historical air route based on the feature matrix of uniform size through the prediction layer to obtain predicted weather; determining the target air route in the historical air routes based on the predicted weather through the output layer; determining a route loss value based on a difference between a result of the target air route in the historical air routes and a label of the target air route in the historical air routes; updating parameters of the air route selection model according to the route loss value.

2. The method of claim 1, wherein, determining a corresponding airspace height according to the model of the target unmanned aerial vehicle comprises: determining a corresponding airspace height according to a pre-prepared airspace division rule and the model of the target unmanned aerial vehicle.

3. The method of claim 1, wherein, determining a set of air routes in the airspace height based on the take-off location information and the destination information comprises: determining a battery power supply station in each air route based on the take-off location information and the destination information; composing air routes with a number of battery power supply stations less than a preset threshold into the set of air routes in the airspace height.

4. An unmanned aerial vehicle route planning apparatus, characterized by, The method comprises the following steps: a first information obtaining module is configured to obtain take-off request information of a target unmanned aerial vehicle, the take-off request information comprising a model of the target unmanned aerial vehicle, take-off location information and destination information; a height determining module is configured to determine a corresponding airspace height according to the model of the target unmanned aerial vehicle; a set determining module is configured to determine a set of air routes in the airspace height based on the take-off location information and the destination information; a second information obtaining module is configured to obtain an occupancy rate, a route distance and weather information of each air route in the set of air routes; a candidate air route determining module is configured to determine a candidate air route of the target unmanned aerial vehicle according to the occupancy rate and the route distance of each air route; a target air route output module is configured to input the candidate air route and corresponding weather information into an air route selection model to obtain a target air route output by the air route selection model; The route selection model is trained by taking historical routes and corresponding historical meteorological information as training samples and taking target routes in the historical routes as training labels. The device further comprises a model training module, which comprises an input layer, a convolutional layer, a pooling layer, a prediction layer, and an output layer. The model training module is configured to acquire historical routes and corresponding historical meteorological information through the input layer. The convolutional layer is configured to perform feature extraction on the historical routes and corresponding historical meteorological information to obtain a feature matrix. The pooling layer is configured to perform pooling on the feature matrix to obtain a feature matrix of uniform size. The prediction layer is configured to predict the meteorological conditions of each historical route based on the feature matrix of uniform size to obtain predicted meteorological conditions. The output layer is configured to determine a target route in the historical routes based on the predicted meteorological conditions. A route loss value is determined based on the difference between the determined target route in the historical routes and the target route label in the historical routes. The parameters of the route selection model are updated according to the route loss value.

5. The apparatus of claim 4, wherein, The height determination module comprises: A height determination sub-module configured to determine the corresponding airspace height according to pre-prepared airspace partition rules and the model of the target UAV.

6. The apparatus of claim 5, wherein, The set determination module comprises: A first set determination sub-module configured to determine battery power supply stations in each route based on the takeoff location information and the destination information. A second set determination sub-module configured to group routes with a number of battery power supply stations less than a preset threshold into a route set in the airspace height. 7.A UAV route planning device, characterized by, The device comprises a memory and a processor. The memory is configured to store a program. The processor is configured to execute the program to implement each step of the UAV route planning method according to any one of claims 1-3.

8. A computer-readable storage medium, characterized in that, The computer program is stored on the computer readable storage medium and is executed by the processor to implement the steps of the UAV route planning method according to any one of claims 1-3.

Images