Decision-making method and decision-making system for low-power autonomous landing based on vertical take-off and landing unmanned aerial vehicle

Abstract

The invention discloses a decision-making method and decision-making system for low-power autonomous landing based on a vertical take-off and landing unmanned aerial vehicle, and relates to the technical field of unmanned aerial vehicle landing decision-making. Comprising the following steps: acquiring topographic point cloud data, establishing a neural network model, and extracting and predicting topographic feature parameters including obstacle density and maximum height; calculating an obstruction influence index and a vertical energy consumption influence index in combination with cruise area point cloud data and real-time environment parameters including environment temperature, wind speed and the like, and correcting a return electric quantity threshold value; dynamically correcting the return electric quantity threshold value by further combining the health parameters of the battery, including storage time, cycle charging times and environment temperature; and comparing the dynamic homeward voyage electric quantity threshold with the real-time electric quantity, and generating a landing or homeward voyage decision instruction. The horizontal flight and vertical take-off and landing energy consumption of the unmanned aerial vehicle under different topographic conditions is comprehensively considered, and correction is performed in combination with the battery performance, so that the electric quantity management strategy of the unmanned aerial vehicle is more intelligent and precise.

Description

Technical Field

[0001] The present invention relates to the technical field of UAV landing decision-making, and specifically to a decision-making method and system for low-power autonomous landing of a vertical take-off and landing UAV. Background Art

[0002] With the rapid development of UAV technology, vertical take-off and landing UAVs (VTOL UAVs) are widely used in fields such as logistics transportation, disaster relief, and topographic mapping due to their efficient cruise capabilities of fixed-wing flight and vertical take-off and landing capabilities of multi-rotor flight. However, when UAVs perform tasks in long distances and complex environments, the limitation of battery endurance becomes a key factor affecting their mission completion rate and safety. Especially during the mission execution, UAVs need to make decisions on returning for landing based on the remaining battery power, and the unplanned crashes caused by insufficient power may cause serious consequences to the UAV itself, mission objectives, and the surrounding environment.

[0003] Existing low-power return or landing decision-making methods mainly rely on the single monitoring of the remaining battery power (SOC, State of Charge) or battery voltage. However, these methods have significant limitations in practical applications. On the one hand, the estimation of SOC or voltage is easily affected by environmental temperature, flight load, and battery aging degree, resulting in misjudgment. For example, high-load flight or low-temperature environment may cause an instantaneous drop in battery voltage, triggering an early return; while battery aging may cause the estimated power value to be inconsistent with the actual remaining capacity, thus delaying the return time. On the other hand, existing methods usually assume that the mission environment is relatively simple and insufficiently consider complex terrain and dynamic environmental factors. In practical applications, the return or landing path of the UAV may pass through areas with dense obstacles, such as high-rise buildings, mountains, or forests, which pose higher requirements for the path planning and energy consumption calculation of the UAV.

[0004] In addition, traditional return power thresholds are usually set statically and fail to dynamically adapt to changes in different environmental parameters. For example, a strong wind environment will significantly increase flight power consumption, and the undulations of complex terrain may lead to higher vertical take-off and landing energy consumption. Static return power thresholds often cannot meet the requirements of these scenarios. This decision-making mode not only limits the flexibility of the UAV but also increases the risk of mission failure due to insufficient estimation of return power.

[0005] Based on this, how to accurately judge the best return or landing time in complex terrain and dynamic environments by combining the battery health status, terrain features, and real-time environmental parameters of the UAV has become a key technical problem in the current low-power management of vertical take-off and landing UAVs.

[0006] In the prior art, the publication number CN117055616A discloses a method for making a low - power autonomous landing decision for a vertical take - off and landing unmanned aerial vehicle. The specific method includes: setting a power return mode and a voltage return mode for the unmanned aerial vehicle; determining whether the power or voltage meets the take - off condition. If it meets, proceed to the next step; if not, terminate the take - off; determine the return mode. If the power return mode is selected, the voltage return mode is also retained, and the percentage of safe landing power is determined. During the flight of the unmanned aerial vehicle, the actual power is obtained in real - time, and whether to return is judged based on the power. When the return condition is triggered, the unmanned aerial vehicle executes the power return decision; if the voltage return mode is selected, the actual voltage is obtained in real - time during the flight of the unmanned aerial vehicle, and whether to return is judged based on the voltage. When the return condition is triggered, the unmanned aerial vehicle executes the voltage return decision; if an error occurs in reading the power data, the voltage return function is automatically activated. However, this method does not mention how to handle obstacles (such as buildings, wires, trees, terrain undulations, etc.) on the return path. During the return process, especially in a low - power state, the unmanned aerial vehicle may not be able to avoid obstacles on the return path, resulting in a collision risk. At the same time, only the influence of wind speed is considered, and the influence of temperature, etc. on power consumption is not considered. Therefore, the accuracy and effectiveness of the judgment and decision - making are reduced due to a single environmental influence.

[0007] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure. Therefore, it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0008] The purpose of the present invention is to provide a decision - making method and a decision - making system for low - power autonomous landing of a vertical take - off and landing unmanned aerial vehicle to solve the problems raised in the above - mentioned background art.

[0009] To achieve the above - mentioned purpose, the present invention provides the following technical solutions:

[0010] A decision - making method for low - power autonomous landing of a vertical take - off and landing unmanned aerial vehicle, the specific steps include:

[0011] Obtain the terrain point cloud data of several terrain feature parameters known as sample point cloud data. Based on the sample point cloud data, establish a neural network model. Use the sample point cloud data as the input of the model and the corresponding terrain feature parameters as labels to train the neural network model to obtain a terrain feature prediction model. The terrain feature parameters include obstacle density and maximum height;

[0012] Collect the terrain point cloud data of the UAV flight cruise area, input the terrain point cloud data of the UAV flight cruise area into the trained terrain feature prediction model, and obtain the predicted values of the terrain feature parameters of the UAV flight cruise area. The predicted values of the terrain feature parameters of the UAV flight cruise area include the predicted value of the obstacle density and the predicted value of the maximum height;

[0013] Obtain the real-time flight altitude data of the UAV during flight cruise. At the same time, obtain the real-time environmental parameters and the horizontal distance from the target takeoff and landing point at the current flight altitude. Calculate the obstruction influence index according to the obtained real-time environmental parameters, the horizontal distance from the takeoff and landing point, and the predicted value of the obstacle density. The environmental parameters include environmental temperature, wind speed, and wind direction angle;

[0014] Obtain the vertical height of the target takeoff and landing point. Through the obtained vertical height of the target takeoff and landing point and the current flight altitude, combined with the total mass of the UAV, calculate and generate the vertical energy consumption influence index. According to the vertical energy consumption influence index, combined with the obstruction influence index, correct the set return power threshold to obtain the influence return power threshold;

[0015] Collect the battery health parameters of the UAV, dynamically correct the influence return power threshold according to the battery health parameters combined with the environmental temperature to obtain the dynamic return power threshold. Compare the dynamic return power threshold with the real-time power of the UAV, and issue corresponding decision instructions according to the comparison result. The battery health parameters include the battery storage time and the number of charge cycles.

[0016] Furthermore, based on the sample point cloud data, establish a neural network model. Among them, establish a terrain feature prediction model based on the long short-term memory network model (LSTM model), select an activation function and an optimization algorithm. Among them, select the Tanh function as the activation function and select Adam as the optimization algorithm for the LSTM model; The formula of the Tanh function is:

[0017]

[0018] In the formula, f(r) represents the Tanh function, and the independent variable r represents the weighted sum of the neuron inputs, that is, the result after the inputs received by the neuron from the previous layer are weighted and summed;

[0019] At the same time, set the hyperparameters of the LSTM model. The hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing quantity, and the number of neurons in the hidden layer;

[0020] Among them, the number of network layers is set to a 3-layer network structure, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch processing quantity is set to 256, and the number of neurons in the hidden layer is 32;

[0021] The input of the terrain feature prediction model that has completed training is terrain point cloud data, and the output is the terrain feature parameters corresponding to the terrain.

[0022] Furthermore, the obstacle impact index is calculated based on the obtained real-time environmental parameters, the horizontal distance from the takeoff and landing point, and the predicted obstacle density. The specific formula for calculating the obstacle impact index is as follows:

[0023]

[0024] In the formula, IES represents the obstacle impact index at the current moment, λ is the obstacle density in the flight cruise area, P wind represents the wind resistance power consumption suffered by the UAV at the current moment, and L represents the horizontal distance between the position of the UAV at the current moment and the takeoff and landing point;

[0025] Among them, the wind resistance power consumption P suffered by the UAV at the current moment wind is calculated through the collected real-time environmental parameters. The specific calculation formula is as follows:

[0026]

[0027] In the formula, ρ is the air density at the current flight altitude, A is the effective windward area of the UAV at the current moment, C d is the air resistance coefficient, and V eff represents the wind speed of the environment where the UAV is located at the current moment at the current flight altitude;

[0028] Among them, the specific formula for calculating the effective windward area of the UAV at the current moment is as follows:

[0029]

[0030] In the formula, θ is the angle between the wind direction angle at the current moment and the flight direction of the UAV, A on and A bise are the projected area in the top view direction and the projected area in the front view direction of the UAV, respectively.

[0031] Furthermore, the vertical energy consumption impact index is calculated by combining the vertical height of the target takeoff and landing point and the current flight altitude obtained, with the total mass of the UAV. The formula for calculating the vertical energy consumption impact index is as follows:

[0032]

[0033] In the formula, E clb is the vertical energy consumption impact index, m is the total mass of the UAV, g is the acceleration due to gravity, h is the height of the target takeoff and landing point, and h f is the flight altitude of the UAV at the current moment, and hfor is the predicted maximum height of the obstacle, and η is the efficiency of the UAV rotor motor;

[0034] Among them, the efficiency η of the UAV rotor motor is corrected by the environmental temperature and the current flight altitude. The specific formula based on which the efficiency η of the UAV rotor motor is calculated is:

[0035]

[0036] In the formula, η0 is the standard efficiency of the UAV rotor motor, T0 is the reference temperature, T hf is the temperature at the position where the UAV is located at the current moment, and β is the altitude correction factor.

[0037] Furthermore, according to the vertical energy consumption impact index, the set return power threshold is corrected in combination with the obstacle impact index to obtain the impact return power threshold. The formula based on which the impact return power threshold is calculated is:

[0038]

[0039] In the formula, yz represents the impact return power threshold at the current moment, IES is the obstacle impact index at the current moment during the UAV flight cruise, yz0 is the initial set impact return power threshold, ω1 and ω2 are the weight coefficients of the obstacle impact index and the vertical energy consumption impact index respectively, where ω1 < ω2 and both ω1 and ω2 are greater than 0.

[0040] Furthermore, based on the battery health parameters, the impact return power threshold is dynamically corrected to obtain the dynamic return power threshold. The formula based on which the dynamic return power threshold is calculated is:

[0041]

[0042] In the formula, yz′ is the dynamic return power threshold, cs and cj are the battery storage time and the number of cyclic charges respectively;

[0043] The dynamic return power threshold is compared with the real-time power of the UAV. According to the comparison result, corresponding decision instructions are issued. The specific logic based on which the corresponding decision instructions are issued is:

[0044] When ZH ≥ 1.0 * yz′, it is judged that the cruise state is healthy, indicating that sufficient power is currently provided for cruising;

[0045] When 0.4 * yz′ ≤ ZH < 1.0 * yz′, it is judged that the cruise state is medium, and a prompt is issued to pay attention to the remaining power;

[0046] When 0 ≤ ZH < 0.4 * yz′, it is judged that the cruise state is low, indicating that a return should be made immediately at present;

[0047] where ZH is the battery display power of the UAV at the current moment.

[0048] The present invention also provides a decision-making system for low-power autonomous landing based on a vertical takeoff and landing UAV. The decision-making system for low-power autonomous landing based on a vertical takeoff and landing UAV is used to execute the above-mentioned decision-making method for low-power autonomous landing based on a vertical takeoff and landing UAV, and includes:

[0049] A prediction model training module, configured to obtain terrain point cloud data of a number of terrain feature parameters as sample point cloud data, establish a neural network model based on the sample point cloud data, use the sample point cloud data as the input of the model, and use the corresponding terrain feature parameters as labels to train the neural network model to obtain a terrain feature prediction model. The terrain feature parameters include obstacle density and maximum height;

[0050] A cruise altitude confirmation module, configured to collect terrain point cloud data of the UAV flight cruise area, input the terrain point cloud data of the UAV flight cruise area into the trained terrain feature prediction model to obtain predicted values of terrain feature parameters of the UAV flight cruise area. The predicted values of terrain feature parameters of the UAV flight cruise area include predicted obstacle density values and maximum height predicted values;

[0051] An environmental disturbance analysis module, configured to obtain the real-time flight altitude data of the UAV during flight cruise, and at the same time obtain the real-time environmental parameters at the current flight altitude and the horizontal distance from the target takeoff and landing point, and calculate an obstacle impact index according to the obtained real-time environmental parameters, the horizontal distance from the takeoff and landing point, and the predicted obstacle density value. The environmental parameters include environmental temperature, wind speed, and wind direction angle;

[0052] A vertical energy consumption analysis module, configured to obtain the vertical height of the target takeoff and landing point, calculate a vertical energy consumption impact index through the obtained vertical height of the target takeoff and landing point and the current flight altitude, and combine the total mass of the UAV. According to the vertical energy consumption impact index, the obstacle impact index is combined to correct the set return power threshold to obtain an impact return power threshold;

[0053] A landing decision generation module, configured to collect the battery health parameters of the UAV, dynamically correct the impact return power threshold according to the battery health parameters and the environmental temperature to obtain a dynamic return power threshold, compare the dynamic return power threshold with the real-time power of the UAV, and issue corresponding decision instructions according to the comparison result. The battery health parameters include battery storage time and number of cycle charges.

[0054] Compared with the prior art, the beneficial effects of the present invention are:

[0055] First, this solution fully considers the terrain complexity of the UAV flight area. By collecting terrain point cloud data and using a neural network model to predict terrain feature parameters, it can accurately obtain core terrain information such as the obstacle density and maximum height in the flight area. During the UAV's cruise flight or return journey, this method effectively avoids the problem that traditional low-battery decision-making methods lack awareness of obstacles in the flight path. By predicting terrain feature parameters, the system can dynamically adjust the flight height and cruise path, thereby reducing the risk of obstacle collisions and improving the safety and mission completion rate of the UAV in complex environments.

[0056] Secondly, real-time environmental parameters are introduced in the calculation of the return battery threshold, including key variables such as environmental temperature, wind speed, and wind direction angle, which can dynamically evaluate the impact of the environment on flight energy consumption. Considering the energy consumption of the UAV's horizontal flight and vertical takeoff and landing under different terrain conditions, the return battery threshold is made more in line with the actual flight requirements. Finally, this solution fully considers the impact of the battery health status on the return decision. Combining health parameters such as battery storage time and number of charge cycles, as well as the dynamic impact of environmental temperature on battery performance, the return battery threshold is further refined. This significantly reduces the return judgment error caused by battery aging or temperature changes, making the UAV's power management strategy more intelligent and precise. Brief Description of the Drawings

[0057] Figure 1 It is a schematic diagram of the overall method flow of the present invention;

[0058] Figure 2 It is a schematic diagram of the overall system structure of the present invention. Detailed Embodiments

[0059] To make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to specific embodiments.

[0060] It should be noted that unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meanings understood by those with ordinary skills in the field to which the present invention belongs. The "first", "second", and similar terms used in the present invention do not indicate any order, quantity, or importance, but are only used to distinguish different components. The terms "including" or "comprising" and the like mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. The terms "connected" or "linked" and the like are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms "up", "down", "left", "right", etc. are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0061] Embodiment:

[0062] Please refer to Figure 1 , the present invention provides a technical solution:

[0063] A decision-making method for low-power autonomous landing based on a vertical takeoff and landing unmanned aerial vehicle, the specific steps include:

[0064] Step 1: Obtain terrain point cloud data of several terrain feature parameters known as sample point cloud data. Based on the sample point cloud data, establish a neural network model. Use the sample point cloud data as the input of the model and the corresponding terrain feature parameters as labels to train the neural network model to obtain a terrain feature prediction model. The terrain feature parameters include obstacle density and maximum height.

[0065] Among them, the obstacle density is specifically the number of obstacle points per unit volume or unit area, and is calculated through the following steps: Divide the terrain point cloud data into several spatial units of fixed size (such as three-dimensional grid volume or two-dimensional grid area); Determine the ground points and obstacle points in the terrain point cloud data, count the number of obstacle point clouds in each spatial unit, regard the number of obstacle point clouds as the number of obstacle points in the unit, and the obstacle density is the ratio of the number of point clouds to the voxel volume or grid area.

[0066] The specific method for determining the ground points and obstacle points in the terrain point cloud data is as follows: Flat ground detection: Use the RANSAC algorithm to perform plane fitting on the point cloud. Randomly extract multiple points in the point cloud data, fit a plane model, calculate the distance from each point to the fitted plane, and classify the points with a distance less than a certain error threshold (such as 0.05 meters) as ground points. Through iterative optimization, find the main plane containing the most points as the ground plane; After removing the ground from the point cloud, the remaining point cloud is basically the obstacle point cloud. According to the height characteristics of the obstacles, mark the points with a height exceeding a certain threshold (such as 0.2 meters) as obstacle points.

[0067] Based on the sample point cloud data, establish a neural network model. Among them, establish a terrain feature prediction model based on the long short-term memory network model LSTM model, select an activation function and an optimization algorithm. Select the Tanh function as the activation function and select Adam as the optimization algorithm for the LSTM model; The formula of the Tanh function is:

[0068]

[0069] In the formula, f(r) represents the Tanh function, and the independent variable r represents the weighted sum of the inputs of the neuron, that is, the result after the inputs received by the neuron from the previous layer are weighted and summed.

[0070] Meanwhile, set the hyperparameters of the LSTM model. The hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing quantity, and the number of neurons in the hidden layer;

[0071] Among them, the number of network layers is set to a 3-layer network structure, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch processing quantity is set to 256, and the number of neurons in the hidden layer is 32;

[0072] Among them, the input of the terrain feature prediction model after training is the terrain point cloud data, and the output is the terrain feature parameters corresponding to the terrain.

[0073] Terrain point cloud data often shows relatively continuous height and density changes. For example, it gradually transitions from flat terrain to steep terrain. This continuous change trend can be regarded as a sequence characteristic, and the LSTM model can effectively capture these trends and predict the distribution density and maximum height of obstacles.

[0074] Terrain point cloud data usually has complex distribution characteristics. It may be necessary to consider both short-term characteristics in local areas and long-term characteristics in a wider range at the same time. In a complex canyon, the obstacle density in some areas may not only depend on neighboring points but also be affected by terrain changes in more distant areas. Similarly, the prediction of the maximum height may involve cumulative height change trends rather than just being based on local height values.

[0075] Terrain point cloud data usually has irregularity. The distribution of points may be uneven in density due to the complexity of the terrain. The LSTM model can, through its sequence processing ability, flexibly adapt to the irregularity of the point cloud data after preprocessing the input data. By organizing the point cloud data into a set of input sequences according to the spatial sequence, the LSTM can effectively learn the feature distribution of the data without relying on a regular grid structure.

[0076] Step 2: Collect the terrain point cloud data of the UAV flight cruise area, and input the terrain point cloud data of the UAV flight cruise area into the trained terrain feature prediction model to obtain the predicted values of the terrain feature parameters of the UAV flight cruise area. The predicted values of the terrain feature parameters of the UAV flight cruise area include the predicted value of the obstacle density and the predicted value of the maximum height.

[0077] The specific steps for obtaining the terrain point cloud data of the UAV flight cruise area include: The UAV is equipped with a sensor device suitable for terrain data collection for collection, such as lidar. Lidar is currently the preferred choice for obtaining high-precision terrain point cloud data. Its working principle is to measure the distance to the ground through laser beams to generate high-resolution three-dimensional point cloud data. Lidar is suitable for complex terrains (such as woods, densely built-up areas, etc.) and can penetrate the tree canopy and restore ground data; or use a high-definition camera, equipped with an RGB high-definition camera, and generate three-dimensional point clouds by aerial photographing multiple overlapping two-dimensional images and using structured light technology or photogrammetry in computer vision. The camera method is suitable for open terrains, but the data restoration effect is poor in areas with dense vegetation.

[0078] After the terrain data is collected, it is necessary to preprocess the original point cloud data or image data, which usually includes the following steps: Synchronize the lidar scan data with the GPS and IMU data of the UAV to ensure the accurate geographical location of the point cloud data. The original point cloud data or image data may contain noise (such as signal reflection errors or external environmental interference), and noise removal needs to be carried out through algorithms (such as the random sample consensus algorithm RANSAC). The processed point cloud data is used as the input of the model.

[0079] Step 3: Obtain the real-time flight altitude data of the UAV during flight cruise, and at the same time obtain the real-time environmental parameters at the current flight altitude and the horizontal distance from the target takeoff and landing point. Calculate the obstruction influence index according to the obtained real-time environmental parameters, the horizontal distance from the takeoff and landing point, and the predicted value of the obstacle density. The environmental parameters include environmental temperature, wind speed, and wind direction angle.

[0080] Calculate the obstruction influence index according to the obtained real-time environmental parameters, the horizontal distance from the takeoff and landing point, and the predicted value of the obstacle density. The specific formula based on which the obstruction influence index is calculated is:

[0081]

[0082] In the formula, IES represents the obstruction influence index at the current moment, λ is the obstacle density in the flight cruise area, P wind represents the wind resistance power consumption suffered by the UAV at the current moment, and L represents the horizontal distance between the position of the UAV at the current moment and the takeoff and landing point;

[0083] It should be noted that the obstruction influence index IES at the current moment during the UAV flight cruise is characterized by comprehensively considering the obstacle density in the flight cruise area, the real-time horizontal distance from the takeoff and landing point, and the wind resistance power consumption. Among them, the larger the value of the obstruction influence index IES at the current moment during the UAV flight cruise, the greater the obstruction during the UAV cruise and the more power required.

[0084] Among them, the obstacle density λ in the flight cruise area represents the complexity of the obstacle distribution in the environment, directly affecting the flight path planning and energy consumption of the UAV. When the obstacle density is high, the probability of emergencies increases, and the UAV needs to reserve more power to avoid emergency operations, which will increase the flight path length and energy consumption. Therefore, the obstacle density λ in the flight cruise area is proportional to the current moment's obstacle influence index IES during the UAV's flight cruise. The logarithmic function ln(1 + λ) is used to appropriately smooth the impact of the growth of the obstacle density value on the result. The obstacle density usually changes sharply due to terrain complexity, but the logarithmic transformation can avoid the unreasonable amplification of the current moment's obstacle influence index IES0 during the UAV's flight cruise when the density value is too large.

[0085] The horizontal distance L between the UAV's current position and the takeoff and landing point directly affects the energy and time required for the return flight. As the horizontal distance increases, the return flight path of the UAV becomes longer, and the energy consumption also increases significantly. Therefore, the horizontal distance L between the UAV's current position and the takeoff and landing point is proportional to the obstacle influence index IES, and the square form L 2 reflects the non-linear influence of distance on energy consumption. Energy consumption usually grows non-linearly with flight distance, which can more accurately describe the non-linear relationship between energy consumption and the increase in distance of the return flight path.

[0086] Wind speed and wind direction are important environmental factors affecting the endurance of the UAV. When flying in a direction inconsistent with the wind direction, the UAV needs to overcome higher wind resistance power consumption, so the energy consumption increases significantly. Therefore, the current wind resistance power consumption P of the UAV wind is proportional to the obstacle influence index IES, and through the exponential function in the denominator represents the proportional relationship, converting the wind resistance power consumption value into a non-linear relationship, indicating the significant influence of the current wind resistance power consumption P of the UAV wind on the obstacle influence index IES.

[0087] Among them, the current wind resistance power consumption P of the UAV wind is calculated through the collected real-time environmental parameters. The specific calculation formula is as follows:

[0088]

[0089] In the formula, ρ is the air density at the current flight altitude, A is the effective windward area of the UAV at the current moment, C d is the air resistance coefficient, and V eff represents the wind speed of the environment where the UAV is located at the current moment at the current flight altitude;

[0090] Among them, the air resistance coefficient Cd The acquisition method is as follows: Place the drone or its model in a wind tunnel, measure the aerodynamic force (especially the drag force) data by changing the wind speed, and then calculate the drag coefficient, or refer to the typical air drag coefficient values of similar aircraft, generally between 0.1 and 1.5.

[0091] The formula for specifically calculating the effective windward area of the drone at the current moment is as follows:

[0092]

[0093] In the formula, θ is the angle between the wind direction angle at the current moment and the flight direction of the drone, A on and A bise are the projected area in the top view direction and the projected area in the front view direction of the drone respectively.

[0094] Install a small electronic anemometer (such as an ultrasonic anemometer or a mechanical vane anemometer) on the drone. The anemometer measures the direction and magnitude of the wind speed in real time. The wind speed direction is the wind direction angle. The IMU sensor (including an accelerometer, a gyroscope, and a magnetometer) can obtain the attitude and motion direction of the drone in real time, and determine the angle between the wind direction angle at the current moment and the flight direction of the drone based on the obtained motion direction and wind speed direction.

[0095] The effective windward area is calculated through the average area of the projected area in the top view direction and the projected area in the front view direction of the drone. The projected area in the top view direction specifically refers to the projected area observed from the top of the drone (perpendicular to the ground). This is the horizontal projection of the drone from the top view, mainly composed of the horizontal projections of the fuselage, brackets, propellers, etc. The projected area in the front view direction specifically refers to the projected area observed from the front of the drone (parallel to the flight direction). This is the vertical projection of the drone from the front view, mainly composed of the front projections of the fuselage, brackets, and other components.

[0096] Step 4: Obtain the vertical height of the target takeoff and landing point. Based on the obtained vertical height of the target takeoff and landing point and the current flight height, combined with the total mass of the drone, calculate and generate a vertical energy consumption impact index. According to the vertical energy consumption impact index, combined with the obstruction impact index, correct the set return power threshold to obtain the impact return power threshold.

[0097] Based on the obtained vertical height of the target takeoff and landing point and the current flight height, combined with the total mass of the drone, calculate and generate a vertical energy consumption impact index. The formula for calculating the vertical energy consumption impact index is as follows:

[0098]

[0099] In the formula, E clbis the vertical energy consumption impact index, m is the total mass of the drone, g is the acceleration due to gravity, h is the height of the target takeoff and landing point, h f is the flight height of the drone at the current moment, h for is the predicted maximum height of the obstacle, and η is the efficiency of the drone's rotor motor;

[0100] Among them, the vertical energy consumption impact index represents the energy consumed during vertical movement. Among them, the vertical energy consumption impact index E clb The larger the value, the more power is consumed.

[0101] During vertical flight, the drone needs to overcome gravity to complete climbing, or control the rate through the action of the motor or aerodynamics during descent. When the drone climbs, it needs to consume energy to counteract its own weight (i.e., overcome gravity), and this energy is directly related to the change in gravitational potential energy. Among them, the maximum value of the flight height of the drone at the current moment and the predicted maximum height of the obstacle is taken to calculate the height difference. Taking the maximum value of the two to calculate the difference can ensure that the drone has sufficient power for vertical landing. Therefore, the vertical energy consumption impact index is proportional to the difference between the maximum value of the flight height of the drone at the current moment and the predicted maximum height of the obstacle and the height of the target takeoff and landing point. The greater the mass, the greater the gravity borne by the drone, so the greater the power required for vertical flight. Therefore, the total mass m of the drone and the vertical energy consumption impact index E clb are proportional.

[0102] When the drone's rotor motor converts electrical energy into mechanical energy, it is not 100% efficient, and there is a certain amount of energy loss. The motor efficiency represents the ratio of the electrical energy consumed by the motor to the useful mechanical energy converted. Since energy consumption is inversely proportional to efficiency, is introduced into the formula to correct the consumption of gravitational potential energy to reflect the actual energy consumption.

[0103] Among them, the efficiency η of the drone's rotor motor is obtained by correcting the ambient temperature and the current flight height. The specific formula based on which the efficiency η of the drone's rotor motor is calculated is:

[0104]

[0105] In the formula, η0 is the standard efficiency of the drone's rotor motor, T0 is the reference temperature, T hf is the temperature at the location of the drone at the current moment, and β is the altitude correction factor.

[0106] The deviation of the operating temperature from the reference temperature T0 will affect the operating efficiency of the mechanical and electrical parts. The closer the ambient temperature is to the reference temperature T0, the closer the performance of the motor is to the ideal state. Therefore, the efficiency η of the drone's rotor motor and |T0 - T hfIt is inversely proportional. Generally, the reference temperature is taken as 25°C. The standard efficiency η0 of the UAV rotor motor can be obtained from the product parameter manual provided by the merchant.

[0107] As the altitude increases, the air density decreases, the cooling capacity drops, the heat dissipation effect deteriorates, and the efficiency of the motor decreases. Therefore, the efficiency η of the UAV rotor motor is inversely proportional to the flight altitude h of the UAV at the current moment. f It is inversely proportional.

[0108] Among them, the altitude correction factor β is used to represent the influence of different altitude levels on the efficiency of the UAV rotor motor. It can be set in combination with expert experience, generally taking a value between 0.1 and 0.5.

[0109] According to the vertical energy consumption impact index, the set return power threshold is corrected in combination with the obstacle impact index to obtain the impact return power threshold. The formula based on which the impact return power threshold is calculated is as follows:

[0110]

[0111] In the formula, yz represents the impact return power threshold at the current moment, IES is the obstacle impact index at the current moment during the UAV flight cruise, yz0 is the set initial impact return power threshold, ω1 and ω2 are the weight coefficients of the obstacle impact index and the vertical energy consumption impact index respectively, where ω1 < ω2 and both ω1 and ω2 are greater than 0.

[0112] It should be noted that by comprehensively considering the obstacle impact index and the vertical energy consumption impact index, the impact return power threshold yz at the current moment is characterized. The larger yz is, the more power needs to be reserved for return landing. It has been stated that both the obstacle impact index and the vertical energy consumption impact index are proportional to the energy power consumption. Therefore, both the obstacle impact index and the vertical energy consumption impact index are proportional to the impact return power threshold yz at the current moment.

[0113] In the form of a square root It represents the proportional relationship of the vertical energy consumption impact index to the impact return power threshold. At the same time, the square root indicates a non-linear reduction of it. And the contribution of vertical energy consumption to the return power is important but non-linear because the energy consumption during vertical climb is usually greater than that during horizontal flight. However, during return, the UAV may recover some energy through gravitational potential energy during the descent process (such as some motors supporting energy feedback). Therefore, the impact of vertical energy consumption on the return power is not completely linear.

[0114] In the form of IES squared 2 It represents the additional energy consumption caused by external environmental conditions (such as wind speed, obstacle density, emergencies, etc.) during flight. Using IES squared 2After magnifying it, it shows that the contribution of the hindering factor to the return flight power is non-linear. Especially in harsh environments, its impact may be more significant.

[0115] Since the vertical energy consumption is mainly related to the height difference between the takeoff and landing points and the flight altitude, its value is basically fixed during the flight, and the influence is relatively stable. The hindering influence is dynamic and may increase rapidly due to environmental changes, with a greater potential impact on energy consumption. Therefore, a higher weight needs to be assigned, so ω1 < ω2 and both ω1 and ω2 are greater than 0.

[0116] The initial threshold yz0 of the influencing return flight power can be set according to the actual environment combined with expert experience, generally 10% - 30%.

[0117] Step 5: Collect the battery health parameters of the drone, dynamically correct the threshold of the influencing return flight power according to the battery health parameters combined with the environmental temperature to obtain the dynamic return flight power threshold, compare the dynamic return flight power threshold with the real-time power of the drone, and issue corresponding decision instructions according to the comparison result. The battery health parameters include the battery storage time and the number of charge and discharge cycles.

[0118] Dynamically correct the threshold of the influencing return flight power based on the battery health parameters to obtain the dynamic return flight power threshold. The formula based on which the dynamic return flight power threshold is calculated is:

[0119]

[0120] In the formula, yz′ is the dynamic return flight power threshold, cs and cj are the battery storage time and the number of charge and discharge cycles respectively;

[0121] During the long-term storage of the battery, the capacity will decay due to self-discharge and chemical property changes (such as electrolyte decomposition, lithium plating on the negative electrode, etc.). The longer the storage time, the more obvious the capacity loss of the battery, and the lower the actual available power. For each complete charge and discharge cycle of the battery, chemical reactions will occur in its internal active materials, resulting in irreversible capacity decline. As the number of charge and discharge cycles increases, the total capacity of the battery will gradually decrease, the internal resistance will increase, and the output power will decrease. Therefore, it is necessary to increase the threshold of the influencing return flight power to ensure that there is actually sufficient power for a return landing. Therefore, the dynamic return flight power threshold is proportional to the battery storage time and the number of charge and discharge cycles. By Comprehensively considering the battery storage time and the number of charge and discharge cycles, the square root of the sum of their squares is used to quantify the overall health status of the battery.

[0122] Temperature is an important factor affecting battery performance. A low-temperature environment will significantly affect the output power and capacity of the battery. The ambient temperature at the cruise altitude usually decreases as the flight altitude increases. A low-temperature environment will cause the chemical reaction rate of the battery to decrease and the discharge performance to deteriorate. Therefore, the dynamic return flight power threshold yz′ is inversely proportional to the temperature.

[0123] Among them, the specific battery storage time refers to the number of days from the first use of the UAV battery to the current date, and the specific number of cyclic charges refers to the number of times of charging within the number of days from the first use of the UAV battery to the current date.

[0124] Compare the dynamic return flight power threshold with the real-time power of the UAV. According to the comparison result, issue a corresponding decision instruction. The specific logic for issuing the corresponding decision instruction is as follows:

[0125] When ZH≥1.0*yz′, it is determined that the cruise state is healthy, indicating that sufficient power is currently provided for cruising;

[0126] When 0.4*yz′≤ZH<1.0*yz′, it is determined that the cruise state is medium, and a prompt is issued to pay attention to the remaining power;

[0127] When 0≤ZH<0.4*yz′, it is determined that the cruise state is low, indicating that a return flight should be carried out immediately;

[0128] Among them, ZH is the power displayed by the UAV battery at the current moment.

[0129] Please refer to Figure 2 , the present invention also provides a decision-making system for low-power autonomous landing of a vertical takeoff and landing UAV. The decision-making system for low-power autonomous landing of a vertical takeoff and landing UAV is used to execute the above-mentioned decision-making method for low-power autonomous landing of a vertical takeoff and landing UAV, including:

[0130] A prediction model training module, which is used to obtain terrain point cloud data with several known terrain feature parameters as sample point cloud data, establish a neural network model based on the sample point cloud data, use the sample point cloud data as the input of the model, and use the corresponding terrain feature parameters as labels to train the neural network model to obtain a terrain feature prediction model. The terrain feature parameters include obstacle density and maximum height;

[0131] A cruise altitude confirmation module, which is used to collect the terrain point cloud data of the UAV flight cruise area, input the terrain point cloud data of the UAV flight cruise area into the trained terrain feature prediction model, and obtain the predicted values of the terrain feature parameters of the UAV flight cruise area. The predicted values of the terrain feature parameters of the UAV flight cruise area include the predicted value of obstacle density and the predicted value of maximum height;

[0132] An environmental disturbance analysis module, which is used to obtain the real-time flight altitude data of the UAV during flight cruise, and at the same time obtain the real-time environmental parameters at the current flight altitude and the horizontal distance from the target takeoff and landing point. According to the obtained real-time environmental parameters, the horizontal distance from the takeoff and landing point, and the predicted value of obstacle density, the obstacle influence index is calculated. The environmental parameters include environmental temperature, wind speed, and wind direction angle;

[0133] A vertical energy consumption analysis module, which is used to obtain the vertical height of the target takeoff and landing point. By combining the vertical height of the target takeoff and landing point and the current flight altitude, and considering the total mass of the UAV, the vertical energy consumption influence index is calculated. According to the vertical energy consumption influence index, combined with the obstacle influence index, the set return power threshold is corrected to obtain the influence return power threshold;

[0134] A landing decision generation module, which is used to collect the battery health parameters of the UAV, dynamically correct the influence return power threshold according to the battery health parameters combined with the environmental temperature to obtain the dynamic return power threshold, compare the dynamic return power threshold with the real-time power of the UAV, and issue corresponding decision instructions according to the comparison result. The battery health parameters include battery storage time and number of cycle charges.

[0135] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula that is closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0136] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0137] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0138] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, and all of them should be covered within the protection scope of the present application.

Claims

1. A decision method for low-power autonomous landing of a vertical take-off and landing UAV, characterized in that: The specific steps include: Acquire terrain point cloud data with several known terrain feature parameters as sample point cloud data, establish a neural network model based on the sample point cloud data, use the sample point cloud data as the input of the model, and use the corresponding terrain feature parameters as labels to train the neural network model to obtain a terrain feature prediction model, wherein the terrain feature parameters include obstacle density and maximum height; Collecting terrain point cloud data of the UAV flight cruising area, inputting the terrain point cloud data of the UAV flight cruising area into the trained terrain feature prediction model, and obtaining the terrain feature parameter prediction value of the UAV flight cruising area, wherein the terrain feature parameter prediction value of the UAV flight cruising area includes the obstacle density prediction value and the maximum height prediction value; Acquire the real-time flight altitude data of the UAV during flight cruising, and at the same time acquire the real-time environmental parameters at the current flight altitude and the horizontal distance from the target take-off and landing point, and calculate the obstacle impact index according to the obtained real-time environmental parameters, the horizontal distance from the take-off and landing point and the predicted value of obstacle density, wherein the environmental parameters include ambient temperature, wind speed and wind direction angle; Obtain the vertical height of the target take-off and landing point, calculate and generate the vertical energy consumption impact index by combining the vertical height of the target take-off and landing point and the current flight altitude with the total mass of the UAV, and correct the set return power threshold according to the vertical energy consumption impact index and the obstacle impact index to obtain the return power threshold; The battery health parameters of the drone are collected, and the threshold affecting the return power is dynamically corrected according to the battery health parameters and the ambient temperature to obtain a dynamic return power threshold. The dynamic return power threshold is compared with the real-time power of the drone, and corresponding decision instructions are issued according to the comparison results. The battery health parameters include battery storage time and number of charge cycles.

2. The decision method for low-battery autonomous landing of a vertical take-off and landing UAV according to claim 1, characterized in that: Based on the sample point cloud data, a neural network model is established. Among them, a terrain feature prediction model is established based on the long short-term memory network model LSTM model, and an activation function and an optimization algorithm are selected. The Tanh function is selected as the activation function, and Adam is selected as the optimization algorithm of the LSTM model; the formula of the Tanh function is: In the formula, f(r) represents the Tanh function, and the independent variable r represents the weighted sum of the neuron's input, that is, the result of the weighted summation of the input received by the neuron from the previous layer; At the same time, the hyperparameters of the LSTM model are set, and the hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing number, and the number of hidden layer neurons; The number of network layers is set to 3 layers, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch size is set to 256, and the number of hidden layer neurons is 32; The input of the trained terrain feature prediction model is terrain point cloud data, and the output is terrain feature parameters of the corresponding terrain.

3. The decision method for low-battery autonomous landing of a vertical take-off and landing UAV according to claim 2 is characterized in that: The obstacle impact index is calculated based on the obtained real-time environmental parameters, the horizontal distance from the take-off and landing points, and the predicted obstacle density. The specific formula for calculating the obstacle impact index is: Where IES represents the obstacle impact index at the current moment, λ is the obstacle density in the flight cruising area, and P wind It represents the wind resistance power consumption of the UAV at the current moment, and L represents the horizontal distance between the UAV's position and the take-off and landing point at the current moment; The wind resistance power consumption P at the current moment wind The calculation is performed based on the collected real-time environmental parameters, and the specific calculation formula is as follows: Where ρ is the air density at the current flight altitude, A is the effective windward area of ​​the drone at the current moment, and C d is the air resistance coefficient, V eff Indicates the wind speed of the drone's environment at the current flight altitude at the current moment; The formula for calculating the effective windward area of ​​the drone at the current moment is: Where θ is the angle between the current wind direction and the flight direction of the drone, A on and A bise They are the projection area of ​​the drone in the downward direction and the projection area in the front direction respectively.

4. The decision method for low-battery autonomous landing of a vertical take-off and landing UAV according to claim 1, characterized in that: The vertical energy consumption impact index is calculated by combining the vertical height of the target take-off and landing point and the current flight altitude with the total mass of the UAV. The formula for calculating the vertical energy consumption impact index is: In the formula, E clb is the vertical energy consumption impact index, m is the total mass of the UAV, g is the acceleration of gravity, h is the target take-off and landing point height, and h f is the current flying height of the drone, h for is the predicted value of the maximum obstacle height, η is the efficiency of the UAV rotor motor; The efficiency η of the UAV rotor motor is obtained by correcting the ambient temperature and the current flight altitude. The specific formula for calculating the efficiency η of the UAV rotor motor is: Where η0 is the standard efficiency of the UAV rotor motor, T0 is the reference temperature, T hf is the temperature of the current position of the UAV, and β is the altitude correction factor.

5. The decision method for low-battery autonomous landing of a vertical take-off and landing UAV according to claim 4, characterized in that: According to the vertical energy consumption impact index and the obstacle impact index, the set return power threshold is corrected to obtain the return power threshold. The formula for calculating the return power threshold is: Where yz represents the current threshold of the power consumption that affects the return trip, IES is the current obstacle impact index during the UAV's flight cruise, yz0 is the set initial threshold of the power consumption that affects the return trip, ω1 and ω2 are the weight coefficients of the obstacle impact index and the vertical energy consumption impact index, respectively, where ω1<ω2 and ω1 and ω2 are both greater than 0.

6. The decision method for low-battery autonomous landing of a vertical take-off and landing UAV according to claim 5, characterized in that: Based on the battery health parameters, the threshold affecting the return power is dynamically corrected to obtain the dynamic return power threshold. The formula for calculating the dynamic return power threshold is: Where yz′ is the dynamic return power threshold, cs and cj are the battery storage time and the number of cycle charging respectively; The dynamic return power threshold is compared with the real-time power of the drone, and the corresponding decision instruction is issued according to the comparison result. The specific logic for issuing the corresponding decision instruction is as follows: When ZH≥1.0*yz′, the cruise status is judged to be healthy, indicating that sufficient power is currently provided for cruising; When 0.4*yz′≤ZH<1.0*yz′, the cruise state is judged to be medium, and a prompt is issued to pay attention to the remaining power; When 0≤ZH<0.4*yz′, the cruise state is judged to be low, indicating that the return should be made immediately; Among them, ZH is the battery power displayed by the drone at the current moment.

7. A decision system for low-power autonomous landing of a vertical take-off and landing UAV, characterized by: The decision system for low-battery autonomous landing of a vertical take-off and landing UAV is used to execute the decision method for low-battery autonomous landing of a vertical take-off and landing UAV according to any one of claims 1 to 6, comprising: A prediction model training module is used to obtain terrain point cloud data with several known terrain feature parameters as sample point cloud data, establish a neural network model based on the sample point cloud data, use the sample point cloud data as the input of the model, and use the corresponding terrain feature parameters as labels to train the neural network model to obtain a terrain feature prediction model, wherein the terrain feature parameters include obstacle density and maximum height; A cruising altitude confirmation module is used to collect terrain point cloud data of the UAV's cruising area, input the terrain point cloud data of the UAV's cruising area into the trained terrain feature prediction model, and obtain the terrain feature parameter prediction value of the UAV's cruising area, wherein the terrain feature parameter prediction value of the UAV's cruising area includes an obstacle density prediction value and a maximum altitude prediction value; The environmental disturbance analysis module is used to obtain the real-time flight altitude data of the UAV during flight and cruise, and at the same time obtain the real-time environmental parameters at the current flight altitude and the horizontal distance from the target take-off and landing point, and calculate the obstacle impact index based on the obtained real-time environmental parameters, the horizontal distance from the take-off and landing point and the obstacle density prediction value, wherein the environmental parameters include ambient temperature, wind speed and wind direction angle; The vertical energy consumption analysis module is used to obtain the vertical height of the target take-off and landing point. The vertical height of the target take-off and landing point and the current flight altitude are combined with the total mass of the UAV to calculate and generate the vertical energy consumption impact index. According to the vertical energy consumption impact index and the obstacle impact index, the set return power threshold is corrected to obtain the return power threshold; The landing decision generation module is used to collect the battery health parameters of the drone, dynamically correct the threshold value that affects the return power according to the battery health parameters and the ambient temperature, obtain the dynamic return power threshold, compare the dynamic return power threshold with the real-time power of the drone, and issue corresponding decision instructions based on the comparison results. The battery health parameters include battery storage time and number of charge cycles.

Images