Dynamic landing method and system of unmanned aerial vehicle

Abstract

The invention discloses a dynamic landing method and system of an unmanned aerial vehicle, and relates to the technical field of unmanned aerial vehicles, when a landing position is received, state information and environment information of the unmanned aerial vehicle are obtained in real time, safety evaluation is performed on the state information and the environment information to obtain a safety state value, whether the safety state value is greater than a preset safety threshold is judged, and if yes, the unmanned aerial vehicle is landed. If the safety state value is smaller than or equal to a safety threshold value, the unmanned aerial vehicle is controlled to land to a landing position with a preset first control strategy according to the state information and the environment information, and if the safety state value is larger than the safety threshold value, the unmanned aerial vehicle is controlled to land to the landing position with a preset second control strategy according to the state information and the environment information. The technical problems that most of existing unmanned aerial vehicle landing control methods plan the landing path of the unmanned aerial vehicle based on a global path planning algorithm, when dynamic obstacles exist in the landing process, the landing path cannot be adjusted in time, and the operation safety of the unmanned aerial vehicle is reduced are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles, and in particular to a dynamic landing method and system for unmanned aerial vehicles. Background Art

[0002] With the rapid development of drone technology, drones are increasingly used in logistics, agriculture, emergency rescue, military reconnaissance and other fields. However, drones face many challenges during landing, especially in complex environments (such as cities, forests and mountains). How to achieve safe landing of drones has become a key issue that needs to be solved urgently.

[0003] At present, most of the existing UAV landing control methods plan the landing path of the UAV based on global path planning algorithms such as the A* algorithm and the Dijkstra algorithm. When there are dynamic obstacles during the landing process, the landing path cannot be adjusted in time, which reduces the safety of the UAV operation. Summary of the invention

[0004] The present invention provides a dynamic landing method and system for an unmanned aerial vehicle, which solves the technical problem that most existing unmanned aerial vehicle landing control methods plan the landing path of the unmanned aerial vehicle based on global path planning algorithms such as the A* algorithm and the Dijkstra algorithm, and when there are dynamic obstacles during the landing process, the landing path cannot be adjusted in time, thereby reducing the safety of the unmanned aerial vehicle operation.

[0005] A first aspect of the present invention provides a dynamic landing method for an unmanned aerial vehicle, comprising:

[0006] When the landing location is received, the status and environment information of the drone is obtained in real time;

[0007] Performing a security assessment on the state information and the environment information to obtain a security state value;

[0008] Determine whether the safety status value is greater than a preset safety threshold;

[0009] If the safety state value is less than or equal to the safety threshold, controlling the UAV to land at the landing position according to the state information and the environmental information using a preset first control strategy;

[0010] If the safety state value is greater than the safety threshold, the UAV is controlled to land at the landing position according to a preset second control strategy based on the state information and the environmental information.

[0011] Optionally, the state information includes the position, flight speed, attitude angle deviation and angular velocity of the drone, the environmental information includes the power of the drone and the positions of multiple obstacles, and the step of performing safety assessment on the state information and the environmental information to obtain a safety state value includes:

[0012] Calculate the distance between the position of the UAV and each obstacle position respectively, and select the minimum distance value from each distance value as the target distance value;

[0013] Performing difference processing on the target distance value and the preset safety distance value to obtain a first evaluation value;

[0014] Performing difference processing on the preset maximum allowable flight speed and the flight speed to obtain a second evaluation value;

[0015] Performing difference processing on a preset maximum attitude angle deviation and the attitude angle deviation to obtain a third evaluation value;

[0016] Performing difference processing on the preset maximum angular velocity and the angular velocity to obtain a fourth evaluation value;

[0017] Performing difference processing on the preset power threshold and the power of the drone to obtain a fifth evaluation value;

[0018] The first evaluation value, the second evaluation value, the third evaluation value, the fourth evaluation value and the fifth evaluation value are weightedly calculated according to a preset safety evaluation weight to obtain a safety status value.

[0019] Optionally, the step of controlling the UAV to land at the landing position according to the state information and the environmental information using a preset first control strategy includes:

[0020] Generate a first landing path according to the drone position and the landing position of the state information;

[0021] Optimizing the first landing path based on a preset local path optimization algorithm, the state information, and the environmental information to obtain a first optimized path;

[0022] Controlling the UAV to land at the landing position along the first optimized path, and acquiring the flight distance of the UAV in real time;

[0023] Determining whether the flight distance is greater than a preset flight threshold;

[0024] When the flight distance is greater than the flight threshold, the process jumps to executing the step of acquiring the status information and environmental information of the UAV in real time.

[0025] Optionally, the step of controlling the UAV to land at the landing position according to the state information and the environmental information using a preset second control strategy includes:

[0026] generating a second landing path according to the drone position and the landing position of the state information;

[0027] Determining a local endpoint position of the drone based on the drone detection distance of the state information and the second landing path;

[0028] Inputting the state information, the environment information and the local end point position into a pre-trained landing path planning model to obtain a second optimized path;

[0029] Determining whether the local end position is the landing position;

[0030] If the local end position is the landing position, controlling the UAV to land to the landing position along the second optimized path;

[0031] If the local end position is not the landing position, controlling the UAV to fly along the second optimized path to the local end position;

[0032] When the UAV reaches the local end position, the process jumps to the step of acquiring the state information and environment information of the UAV in real time.

[0033] Optionally, the training process of the landing path planning model is specifically as follows:

[0034] Acquiring landing training data, and using the landing training data to train a preset initial landing path planning model to obtain training landing path data;

[0035] Calculate the training loss function value and reward function value of the landing training data according to the training landing path data;

[0036] Performing weighted calculation on the training loss function value and the reward function value according to a preset training weight to obtain a target loss value;

[0037] When the target loss value is greater than or equal to the preset standard loss value, the gradient descent method is used to adjust the network parameters of the initial landing path planning model, and the step of training the preset initial landing path planning model using the landing training data to obtain the training landing path data is executed until the target loss value is less than the standard loss value;

[0038] When the target loss value is less than the standard loss value, a landing path planning model is generated.

[0039] Optionally, the landing path planning model is a RNN neural network model.

[0040] A second aspect of the present invention provides a dynamic landing system for an unmanned aerial vehicle, comprising:

[0041] The response module is used to obtain the status information and environmental information of the UAV in real time when receiving the landing position;

[0042] An evaluation module, used for performing a security evaluation on the state information and the environment information to obtain a security state value;

[0043] An analysis module, used to determine whether the safety status value is greater than a preset safety threshold;

[0044] A first control module, configured to control the UAV to land at the landing position according to a preset first control strategy based on the state information and the environmental information if the safety state value is less than or equal to the safety threshold;

[0045] The second control module is used to control the UAV to land at the landing position according to a preset second control strategy based on the state information and the environmental information if the safety state value is greater than the safety threshold.

[0046] A third aspect of the present invention provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the dynamic landing method of a drone as described in any one of the above items.

[0047] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed, the dynamic landing method of the drone as described in any one of the above items is implemented.

[0048] A fifth aspect of the present invention provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer executes the dynamic landing method of the drone as described in any one of the above items.

[0049] It can be seen from the above technical solutions that the present invention has the following advantages:

[0050] The present invention obtains the state information and environmental information of the drone in real time, and performs a safety assessment on the state information and environmental information to obtain a safety state value, and selects a corresponding control strategy to plan the landing path of the drone according to the size of the safety state value, thereby improving the efficiency of the drone landing path planning. It overcomes the technical problem that most of the existing drone landing control methods plan the landing path of the drone based on global path planning algorithms such as the A* algorithm and the Dijkstra algorithm, and when there are dynamic obstacles during the landing process, the landing path cannot be adjusted in time, which reduces the safety of the drone operation. Compared with the traditional landing control method, the present invention calculates the safety state value of the drone in real time, and dynamically selects a suitable control strategy to plan the drone path according to the safety state value, thereby improving the safety of the drone operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0052] Figure 1 A flowchart of a method for dynamic landing of a drone provided in Embodiment 1 of the present invention;

[0053] Figure 2 A flowchart of a method for dynamic landing of a drone provided in Embodiment 2 of the present invention;

[0054] Figure 3 A structural block diagram of a dynamic landing system for a drone provided in Embodiment 3 of the present invention;

[0055] Figure 4 This is a structural block diagram of an electronic device provided in Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0056] The embodiment of the present invention provides a dynamic landing method and system for a drone, which is used to solve the technical problem that most of the existing drone landing control methods plan the landing path of the drone based on global path planning algorithms such as the A* algorithm and the Dijkstra algorithm. When there are dynamic obstacles during the landing process, the landing path cannot be adjusted in time, which reduces the safety of the drone operation.

[0057] In order to make the purpose, features and advantages of the present invention more obvious and easy to understand, the technical solutions 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 embodiments described below are only 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.

[0058] See also Figure 1 , Figure 1 A flowchart of a method for dynamic landing of a drone provided in Embodiment 1 of the present invention.

[0059] The present invention provides a dynamic landing method for an unmanned aerial vehicle, comprising:

[0060] Step 101: When the landing position is received, the state information and environment information of the UAV are obtained in real time;

[0061] In an embodiment of the present invention, when the drone receives the landing position, the state information and environmental information of the drone is obtained through the multimodal perception system on the drone.

[0062] It should be noted that the status information includes the drone's position, flight speed, attitude angle deviation and angular velocity, and the environmental information includes the drone's battery level and the locations of multiple obstacles.

[0063] Step 102: Perform security assessment on the state information and environment information to obtain a security state value;

[0064] In the embodiment of the present invention, the state information and the environment information are input into a preset safety assessment function to obtain a safety state value.

[0065] It should be noted that the security assessment function is specifically:

[0066]

[0067] in, is the safe state value, is the target distance value, is the safety distance value, is the maximum permissible flight speed, is the flight speed, is the maximum attitude angle deviation, is the attitude angle deviation, is the maximum angular velocity, is the angular velocity, is the power threshold, To power the drone, is the first safety assessment weight coefficient, is the second safety assessment weight coefficient, is the third safety assessment weight coefficient, is the fourth safety assessment weight coefficient, It is the fifth safety assessment weight coefficient.

[0068] Step 103: determine whether the safety status value is greater than a preset safety threshold;

[0069] In the embodiment of the present invention, it is determined whether the safety status value is greater than a preset safety threshold.

[0070] Step 104: If the safety state value is less than or equal to the safety threshold, the UAV is controlled to land at the landing position according to the state information and the environment information using a preset first control strategy;

[0071] In an embodiment of the present invention, when the safety status value is less than or equal to the safety threshold, a first landing path is determined based on the UAV position and landing position of the status information, the first landing path is optimized according to a preset local path optimization algorithm, the status information and the environmental information to obtain a first optimized path, the UAV is controlled to land to the landing position along the first optimized path, and the flight distance of the UAV is obtained in real time. When the flight distance is greater than the flight threshold, the process jumps to execute step 101.

[0072] Step 105: If the safety status value is greater than the safety threshold, the UAV is controlled to land at the landing position according to the preset second control strategy based on the status information and the environmental information.

[0073] In an embodiment of the present invention, when the safety status value is greater than the safety threshold, a second landing path is determined based on the drone position and landing position of the status information, and a local terminal position of the drone is determined based on the drone detection distance and the second landing path of the status information. The status information, environmental information and local terminal position are input into a pre-trained landing path planning model to obtain a second optimized path, and it is determined whether the local terminal position is the landing position. If the local terminal position is the landing position, the drone is controlled to land to the landing position along the second optimized path. If the local terminal position is not the landing position, the drone is controlled to fly to the local terminal position along the second optimized path, and the process jumps to execute step 101.

[0074] In an embodiment of the present invention, the state information and environmental information of the drone are acquired in real time, and a safety assessment is performed on the state information and environmental information to obtain a safety state value, and the corresponding control strategy is selected according to the size of the safety state value to plan the landing path of the drone, thereby improving the efficiency of the landing path planning of the drone. This overcomes the technical problem that most of the existing drone landing control methods plan the landing path of the drone based on global path planning algorithms such as the A* algorithm and the Dijkstra algorithm, and when there are dynamic obstacles during the landing process, the landing path cannot be adjusted in time, which reduces the safety of the drone operation. Compared with the traditional landing control method, the present invention calculates the safety state value of the drone in real time, and dynamically selects a suitable control strategy to plan the drone path according to the safety state value, thereby improving the safety of the drone operation.

[0075] See also Figure 2 , Figure 2 A flowchart of a method for dynamic landing of a drone provided in Embodiment 2 of the present invention.

[0076] The present invention provides a dynamic landing method for an unmanned aerial vehicle, comprising:

[0077] Step 201: When the landing position is received, the state information and environment information of the UAV are obtained in real time;

[0078] In an embodiment of the present invention, when the landing position is received, the state information and environmental information of the UAV are acquired in real time through the multimodal perception system.

[0079] It should be noted that the multimodal perception system includes a four-spectrum polarization imaging unit and an electromagnetic field compensation module, which obtains the initial drone position, flight speed, attitude angle deviation, angular velocity, drone power, electromagnetic field change rate and multiple obstacle positions of the drone through the four-spectrum polarization imaging unit. The electromagnetic field compensation module is used to input the initial drone position and electromagnetic field change rate into a preset electromagnetic field compensation function to obtain the drone position.

[0080] The electromagnetic field compensation function is specifically:

[0081]

[0082] in, is the drone location, is the initial UAV position, is the current time, is the initial time, is the compensation amount, is the electromagnetic compensation gain coefficient, is the rate of change of electromagnetic field, For electromagnetic field.

[0083] Step 202: Perform security assessment on the state information and environment information to obtain a security state value;

[0084] Furthermore, the state information includes the position, flight speed, attitude angle deviation and angular velocity of the drone, and the environmental information includes the battery level of the drone and the positions of multiple obstacles. Step 202 includes the following sub-steps:

[0085] S11, respectively calculating the distance values ​​between the position of the UAV and the position of each obstacle, and selecting the minimum distance value from each distance value as the target distance value;

[0086] In the embodiment of the present invention, the distance values ​​between the position of the drone and the position of each obstacle are calculated respectively, and the minimum distance value is selected from all the distance values ​​as the target distance value.

[0087] S12, performing difference processing on the target distance value and the preset safety distance value to obtain a first evaluation value;

[0088] In the embodiment of the present invention, the difference between the target distance value and the preset safety distance value is calculated to obtain the first evaluation value.

[0089] It should be noted that the preset safety distance value is determined according to the size of the drone, the flight speed and the complexity of the operating environment.

[0090] S13, performing difference processing between the preset maximum allowable flight speed and the flight speed to obtain a second evaluation value;

[0091] In the embodiment of the present invention, the difference between the preset maximum allowable flight speed and the flight speed is calculated to obtain the second evaluation value, wherein the maximum allowable flight speed is determined by the hardware performance and design requirements of the drone.

[0092] S14, performing difference processing on the preset maximum attitude angle deviation and the attitude angle deviation to obtain a third evaluation value;

[0093] In the embodiment of the present invention, the difference between the preset maximum attitude angle deviation and the attitude angle deviation is calculated to obtain the third evaluation value.

[0094] S15, performing difference processing between the preset maximum angular velocity and the angular velocity to obtain a fourth evaluation value;

[0095] In the embodiment of the present invention, the difference between the preset maximum angular velocity and the angular velocity is calculated to obtain the fourth evaluation value.

[0096] S16, performing difference processing between the preset power threshold and the power of the drone to obtain a fifth evaluation value;

[0097] In the embodiment of the present invention, the difference between the preset power threshold and the power of the drone is calculated to obtain the fifth evaluation value.

[0098] It should be noted that the power threshold refers to the minimum safe value of the drone battery power.

[0099] S17. Perform weighted calculation on the first evaluation value, the second evaluation value, the third evaluation value, the fourth evaluation value and the fifth evaluation value according to preset safety evaluation weights to obtain a safety status value.

[0100] In an embodiment of the present invention, the first evaluation value, the second evaluation value, the third evaluation value, the fourth evaluation value and the fifth evaluation value are weighted and calculated according to the preset full evaluation weight to obtain the safety state value. The safety evaluation weight includes the first safety evaluation weight coefficient, the second safety evaluation weight coefficient, the third safety evaluation weight coefficient, the fourth safety evaluation weight coefficient and the fifth safety evaluation weight coefficient.

[0101] Step 203: determine whether the safety status value is greater than a preset safety threshold;

[0102] In the embodiment of the present invention, it is determined whether the safety status value is greater than a preset safety threshold.

[0103] Step 204: If the safety status value is less than or equal to the safety threshold, a first landing path is generated according to the drone position and landing position of the status information;

[0104] In the embodiment of the present invention, when the safety state value is less than or equal to the safety threshold, a first landing path is generated according to the drone position and landing position of the state information. The first landing path is a straight line path from the drone position to the landing position (regardless of whether it passes through obstacles).

[0105] Step 205: Optimize the first landing path based on a preset local path optimization algorithm, state information, and environmental information to obtain a first optimized path;

[0106] In an embodiment of the present invention, based on a preset local path optimization algorithm, the first landing path is optimized according to the state information and the environment information to obtain a first optimized path. The local path optimization algorithm includes a dynamic window method (DWA), an artificial potential field method (APF), a rapid exploration random tree (RRT) series, a probabilistic roadmap (PRM), a model predictive control (MPC), an APF+RRT / MPC* method, and a DRL+traditional planning.

[0107] Step 206: Control the UAV to land along the first optimized path to the landing position, and obtain the flight distance of the UAV in real time;

[0108] In an embodiment of the present invention, the UAV is controlled to land at a landing position according to the first optimized path, and the flight distance of the UAV is obtained in real time.

[0109] Step 207: determine whether the flight distance is greater than a preset flight threshold;

[0110] In the embodiment of the present invention, it is determined whether the flight distance is greater than a preset flight threshold.

[0111] Step 208: When the flight distance is greater than the flight threshold, the process jumps to the step of obtaining the status information and environment information of the UAV in real time.

[0112] In the embodiment of the present invention, when the flight distance is greater than the flight threshold, the process jumps to step 201 .

[0113] Step 209: If the safety status value is greater than the safety threshold, the UAV is controlled to land at the landing position according to the preset second control strategy based on the status information and the environmental information.

[0114] Further, step 209 includes the following sub-steps:

[0115] S21, generating a second landing path according to the drone position and landing position of the status information;

[0116] In an embodiment of the present invention, a second landing path is generated according to the drone position and landing position of the state information, wherein the second landing path is a straight line path from the drone position to the landing position (regardless of whether it passes through obstacles).

[0117] S22, determining the local end point position of the UAV based on the UAV detection distance and the second landing path of the state information;

[0118] In the embodiment of the present invention, 1. The drone detection distance of the status information is used as the selected target distance. 2. According to the selected target distance, the point on the second landing path that is farthest from the current drone is selected as the local end point position. 3. If the local end point position coincides with the obstacle position, the selected target distance is reduced by half as the new selected target distance. 4. Step 2 is re-executed until the selected local end point position does not coincide with the obstacle position.

[0119] It should be noted that the local end points are all located on the second landing path.

[0120] S23, inputting the state information, environmental information and local terminal position into a pre-trained landing path planning model to obtain a second optimized path;

[0121] In an embodiment of the present invention, the state information, the environment information and the local end point position are used as inputs of a pre-trained landing path planning model to obtain a second optimized path.

[0122] It should be noted that the landing path planning model is an RNN neural network model.

[0123] It should be noted that the training process of the landing path planning model is as follows:

[0124] A1. Acquire landing training data, and use the landing training data to train a preset initial landing path planning model to obtain training landing path data;

[0125] Landing training data refers to the historical trajectory data of the drone landing.

[0126] In an embodiment of the present invention, historical trajectory data of the UAV landing is obtained, and a preset initial landing path planning model is trained using the historical trajectory data to obtain training landing path data.

[0127] A2. Calculate the training loss function value and reward function value of the landing training data according to the training landing path data;

[0128] In an embodiment of the present invention, the training drop path data and the landing training data are input into a preset loss function to obtain a training loss function value. The training drop path data and the landing training data are input into a preset reward function to obtain a reward function value. The reward function value is the sum of the reward values ​​at each moment.

[0129] It should be noted that the loss function is specifically:

[0130]

[0131] in, is the training loss function value, is the sample size, is the action of the ith expert, is the ith predicted action, is the state of the ith expert, is the state predicted for the i-th sample, where i is the sample number.

[0132] It should be noted that the reward function is specifically:

[0133]

[0134] in, is the reward value at time t, is the position of the UAV at time t, is the target landing position, is the distance between the drone and the obstacle.

[0135] A3. Perform weighted calculation on the training loss function value and the reward function value according to the preset training weight to obtain the target loss value;

[0136] In an embodiment of the present invention, a training loss function value and a reward function value are weightedly calculated according to preset training weights to obtain a target loss value, wherein the training weights include a loss weight coefficient and a reward weight coefficient.

[0137] A4. When the target loss value is greater than or equal to the preset standard loss value, the gradient descent method is used to adjust the network parameters of the initial landing path planning model, and the execution is jumped to the step of training the preset initial landing path planning model with the landing training data to obtain the training landing path data until the target loss value is less than the standard loss value;

[0138] In an embodiment of the present invention, when the target loss value is greater than or equal to a preset standard loss value, the gradient descent method is used to adjust the network parameters of the initial landing path planning model, and the process jumps to execute A1 until the target loss value is less than the standard loss value.

[0139] A5. When the target loss value is less than the standard loss value, a landing path planning model is generated.

[0140] In the embodiment of the present invention, when the target loss value is less than the standard loss value, a landing path planning model is generated.

[0141] S24, determining whether the local end point position is a landing position;

[0142] In the embodiment of the present invention, it is determined whether the local end point position coincides with the landing position.

[0143] S25, if the local end position is the landing position, controlling the UAV to land to the landing position along the second optimized path;

[0144] In an embodiment of the present invention, if the local end point position coincides with the landing position, the UAV is controlled to land at the landing position along the second optimized path.

[0145] S26, if the local end position is not the landing position, controlling the UAV to fly along the second optimized path to the local end position;

[0146] In an embodiment of the present invention, if the local end point position does not coincide with the landing position, the UAV is controlled to fly along the second optimized path to the local end point position.

[0147] S27: When the UAV reaches the local end position, jump to the step of obtaining the status information and environmental information of the UAV in real time.

[0148] In the embodiment of the present invention, when the drone flies to the local end position, the process jumps to step 201 .

[0149] In an embodiment of the present invention, the state information and environmental information of the drone are acquired in real time, and a safety assessment is performed on the state information and environmental information to obtain a safety state value, and the corresponding control strategy is selected according to the size of the safety state value to plan the landing path of the drone, thereby improving the efficiency of the landing path planning of the drone. This overcomes the technical problem that most of the existing drone landing control methods plan the landing path of the drone based on global path planning algorithms such as the A* algorithm and the Dijkstra algorithm, and when there are dynamic obstacles during the landing process, the landing path cannot be adjusted in time, which reduces the safety of the drone operation. Compared with the traditional landing control method, the present invention calculates the safety state value of the drone in real time, and dynamically selects a suitable control strategy to plan the drone path according to the safety state value, thereby improving the safety of the drone operation.

[0150] See also Figure 3 , Figure 3 This is a structural block diagram of a dynamic landing system for an unmanned aerial vehicle provided in Embodiment 3 of the present invention.

[0151] The present invention provides a dynamic landing system for an unmanned aerial vehicle, comprising:

[0152] The response module 301 is used to obtain the status information and environment information of the UAV in real time when receiving the landing position;

[0153] An evaluation module 302 is used to perform a security evaluation on the state information and the environment information to obtain a security state value;

[0154] An analysis module 303 is used to determine whether the safety status value is greater than a preset safety threshold;

[0155] The first control module 304 is used to control the UAV to land to the landing position according to the preset first control strategy according to the state information and the environment information if the safety state value is less than or equal to the safety threshold;

[0156] The second control module 305 is used to control the UAV to land at the landing position according to the preset second control strategy based on the state information and the environmental information if the safety state value is greater than the safety threshold.

[0157] Furthermore, the state information includes the position, flight speed, attitude angle deviation and angular velocity of the drone, and the environmental information includes the power of the drone and the positions of multiple obstacles. The evaluation module 302 includes:

[0158] The distance measurement submodule is used to calculate the distance values ​​between the drone position and each obstacle position, and select the minimum distance value from each distance value as the target distance value;

[0159] An evaluation submodule, used for performing difference processing on the target distance value and the preset safety distance value to obtain a first evaluation value;

[0160] Performing difference processing on the preset maximum allowable flight speed and the flight speed to obtain a second evaluation value;

[0161] Performing difference processing on the preset maximum attitude angle deviation and the attitude angle deviation to obtain a third evaluation value;

[0162] Performing difference processing on the preset maximum angular velocity and the angular velocity to obtain a fourth evaluation value;

[0163] Perform difference processing on the preset power threshold and the power of the drone to obtain a fifth evaluation value;

[0164] The first weighted submodule is used to perform weighted calculation on the first evaluation value, the second evaluation value, the third evaluation value, the fourth evaluation value and the fifth evaluation value according to a preset safety evaluation weight to obtain a safety status value.

[0165] Furthermore, the first control module 304 includes:

[0166] A first landing path submodule, used to generate a first landing path according to the drone position and landing position of the state information;

[0167] A first optimization submodule, configured to optimize the first landing path based on a preset local path optimization algorithm, state information, and environmental information to obtain a first optimized path;

[0168] A first control submodule is used to control the UAV to land along a first optimized path to a landing position, and obtain the flight distance of the UAV in real time;

[0169] The first analysis submodule is used to determine whether the flight distance is greater than a preset flight threshold;

[0170] When the flight distance is greater than the flight threshold, the process jumps to the step of obtaining the status information and environment information of the drone in real time.

[0171] Furthermore, the second control module 305 includes:

[0172] A second landing path submodule, used to generate a second landing path according to the drone position and landing position of the status information;

[0173] A local endpoint submodule, for determining the local endpoint position of the UAV based on the UAV detection distance and the second landing path of the state information;

[0174] A second optimization submodule is used to input the state information, the environment information and the local end point position into a pre-trained landing path planning model to obtain a second optimized path;

[0175] The second analysis submodule is used to determine whether the local end point position is a landing position;

[0176] If the local end point is the landing position, the UAV is controlled to land at the landing position along the second optimized path;

[0177] If the local end point is not the landing position, the UAV is controlled to fly along the second optimized path to the local end point;

[0178] When the UAV reaches the local end position, the execution jumps to the step of obtaining the status information and environmental information of the UAV in real time.

[0179] Furthermore, the training process of the landing path planning model is as follows:

[0180] Acquire landing training data, and use the landing training data to train a preset initial landing path planning model to obtain training landing path data;

[0181] Calculate the training loss function value and reward function value of the landing training data according to the training landing path data;

[0182] The training loss function value and the reward function value are weighted according to the preset training weights to obtain the target loss value;

[0183] When the target loss value is greater than or equal to the preset standard loss value, the gradient descent method is used to adjust the network parameters of the initial landing path planning model, and the step of training the preset initial landing path planning model with the landing training data to obtain the training landing path data is executed until the target loss value is less than the standard loss value;

[0184] When the target loss value is less than the standard loss value, a landing path planning model is generated.

[0185] Furthermore, the landing path planning model is a RNN neural network model.

[0186] See also Figure 4 , Figure 4 This is a structural block diagram of an electronic device provided in Embodiment 4 of the present invention.

[0187] An electronic device according to an embodiment of the present invention includes: a memory 401 and a processor 402, wherein the memory 402 stores a computer program; when the computer program is executed by the processor 402, the processor 402 executes the dynamic landing method of the drone according to any of the above embodiments.

[0188] The memory 401 may be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read-only memory), an EPROM, a hard disk or a ROM. The memory 401 has a storage space 403 for a program code 413 for executing any method step in the above method. For example, the storage space 403 for the program code may include individual program codes 413 for implementing the various steps in the above method, respectively. These program codes may be read from or written to one or more computer program products. These computer program products include program code carriers such as a hard disk, a compact disk (CD), a memory card or a floppy disk. The program code may be compressed, for example, in an appropriate form. When these codes are run by a computing and processing device, the computing and processing device is caused to execute the various steps in the above-described method.

[0189] Embodiment 5 of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the dynamic landing method of the drone as in any of the above embodiments is implemented.

[0190] Embodiment 6 of the present invention further provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer executes the dynamic landing method of a drone as described in any of the above embodiments.

[0191] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0192] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0193] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0194] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0195] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0196] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A dynamic landing method for an unmanned aerial vehicle, characterized in that: include: When the landing location is received, the status and environment information of the drone is obtained in real time; Performing a security assessment on the state information and the environment information to obtain a security state value; Determine whether the safety status value is greater than a preset safety threshold; If the safety state value is less than or equal to the safety threshold, controlling the UAV to land at the landing position according to the state information and the environmental information using a preset first control strategy; If the safety state value is greater than the safety threshold, the UAV is controlled to land at the landing position according to a preset second control strategy based on the state information and the environmental information.

2. The dynamic landing method of a drone according to claim 1, characterized in that: The state information includes the position, flight speed, attitude angle deviation and angular velocity of the drone, the environmental information includes the power of the drone and the positions of multiple obstacles, and the step of performing safety assessment on the state information and the environmental information to obtain a safety state value includes: Calculate the distance between the position of the UAV and each obstacle position respectively, and select the minimum distance value from each distance value as the target distance value; Performing difference processing on the target distance value and the preset safety distance value to obtain a first evaluation value; Performing difference processing on the preset maximum allowable flight speed and the flight speed to obtain a second evaluation value; Performing difference processing on a preset maximum attitude angle deviation and the attitude angle deviation to obtain a third evaluation value; Performing difference processing on the preset maximum angular velocity and the angular velocity to obtain a fourth evaluation value; Performing difference processing on the preset power threshold and the power of the drone to obtain a fifth evaluation value; The first evaluation value, the second evaluation value, the third evaluation value, the fourth evaluation value and the fifth evaluation value are weightedly calculated according to a preset safety evaluation weight to obtain a safety status value.

3. The dynamic landing method of a drone according to claim 1, characterized in that: The step of controlling the UAV to land at the landing position according to the state information and the environmental information using a preset first control strategy comprises: Generate a first landing path according to the drone position and the landing position of the state information; Optimizing the first landing path based on a preset local path optimization algorithm, the state information, and the environmental information to obtain a first optimized path; Controlling the UAV to land at the landing position along the first optimized path, and acquiring the flight distance of the UAV in real time; Determining whether the flight distance is greater than a preset flight threshold; When the flight distance is greater than the flight threshold, the process jumps to executing the step of acquiring the status information and environmental information of the UAV in real time.

4. The dynamic landing method of a drone according to claim 1, characterized in that: The step of controlling the UAV to land at the landing position according to the state information and the environmental information using a preset second control strategy comprises: generating a second landing path according to the drone position and the landing position of the state information; Determining a local endpoint position of the UAV based on the UAV detection distance of the state information and the second landing path; Inputting the state information, the environment information and the local end point position into a pre-trained landing path planning model to obtain a second optimized path; Determining whether the local end point position is the landing position; If the local end position is the landing position, controlling the UAV to land to the landing position along the second optimized path; If the local end position is not the landing position, controlling the UAV to fly along the second optimized path to the local end position; When the UAV reaches the local end position, the process jumps to the step of acquiring the state information and environment information of the UAV in real time.

5. The dynamic landing method of a drone according to claim 4, characterized in that: The training process of the landing path planning model is specifically as follows: Acquiring landing training data, and using the landing training data to train a preset initial landing path planning model to obtain training landing path data; Calculate the training loss function value and reward function value of the landing training data according to the training landing path data; Performing weighted calculation on the training loss function value and the reward function value according to a preset training weight to obtain a target loss value; When the target loss value is greater than or equal to the preset standard loss value, the gradient descent method is used to adjust the network parameters of the initial landing path planning model, and the step of training the preset initial landing path planning model using the landing training data to obtain the training landing path data is executed until the target loss value is less than the standard loss value; When the target loss value is less than the standard loss value, a landing path planning model is generated.

6. The dynamic landing method of a drone according to claim 4, characterized in that: The landing path planning model is a RNN neural network model.

7. A dynamic landing system for an unmanned aerial vehicle, characterized in that: include: The response module is used to obtain the status information and environmental information of the UAV in real time when receiving the landing position; An evaluation module, used for performing a security evaluation on the state information and the environment information to obtain a security state value; An analysis module, used to determine whether the safety status value is greater than a preset safety threshold; A first control module, configured to control the UAV to land at the landing position according to a preset first control strategy based on the state information and the environmental information if the safety state value is less than or equal to the safety threshold; The second control module is used to control the UAV to land at the landing position according to a preset second control strategy based on the state information and the environmental information if the safety state value is greater than the safety threshold.

8. An electronic device, characterized in that: It includes a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor executes the steps of the dynamic landing method of the drone as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the dynamic landing method of the drone as described in any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, wherein the computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer is caused to execute the dynamic landing method of the drone as described in any one of claims 1-6.

Images