Intelligent flying car

Abstract

The application provides a kind of intelligent flying car, it is related to flying car technical field, the intelligent flying car includes: path planning system, for in the movement process of intelligent flying car from starting point to target point, utilize elevation information to plan the path of ground travel and air flight, calculate optimal mode switching point simultaneously, control the mode switching of ground travel to air flight or air flight to ground travel;Autonomous landing system is used to approach target point and receive no positioning signal in the case, utilize the best safe landing point of space-time RGB image and depth image acquisition, utilize final safe landing point to control the safe landing of the flying car.The intelligent flying car of the application has efficient, energy-saving and accurate path planning function, and can realize autonomous landing in the case of no positioning signal.

Description

Technical Field

[0001] This application relates to the field of flying car technology, and in particular to an intelligent flying car. Background Technology

[0002] Currently, intelligent flying cars face the following three problems in practical applications:

[0003] 1. Intelligent flying car control systems are mostly controlled by computer ground stations or remote controllers. They adopt traditional communication direct drive control methods, which cannot form a control system that can autonomously judge output and conduct risk assessment of control. They also have problems with low safety and reliability.

[0004] 2. Due to mode switching limitations, intelligent flying cars cannot effectively adjust themselves according to the terrain, making it difficult to achieve agile, flexible, and real-time path planning strategy adjustments.

[0005] 3. There is a lack of low-cost, accurate, stable and efficient autonomous landing methods for flying cars. Summary of the Invention

[0006] In view of this, this application provides an intelligent flying car to solve the above-mentioned technical problems.

[0007] This application embodiment provides an intelligent flying car, which is equipped with a positioning terminal and two monocular RGB cameras or one monocular RGB camera and one depth camera. The intelligent flying car includes:

[0008] The path planning system is used to plan the path for ground driving and air flight using elevation information during the movement of an intelligent flying car from the starting point to the target point. At the same time, it calculates the optimal mode switching point and controls the mode switching from ground driving to air flight or from air flight to ground driving.

[0009] An autonomous landing system is used to obtain the optimal safe landing point by using simultaneous RGB and depth images in space and time when the intelligent flying car approaches the target point and cannot receive a positioning signal, and then uses the final safe landing point to control the flying car to land safely.

[0010] The intelligent flying car of this application has efficient, energy-saving and accurate path planning functions, and can achieve autonomous landing in the absence of positioning signals. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0012] Figure 1 This is a functional structure diagram of an intelligent flying car provided in an embodiment of this application;

[0013] Figure 2 A functional structure diagram of the path planning system for an intelligent flying car provided in an embodiment of this application;

[0014] Figure 3 A schematic diagram of the BAS algorithm provided in the embodiments of this application;

[0015] Figure 4 A schematic diagram of the virtual height provided in the embodiments of this application;

[0016] Figure 5 A functional structure diagram of the autonomous landing system of an intelligent flying car provided in an embodiment of this application;

[0017] Figure 6 This is a functional structure diagram of the redundant control system of the intelligent flying car provided in the embodiments of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0019] Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0020] First, a brief introduction to the design concept of the embodiments of this application will be given.

[0021] Currently, intelligent flying cars are widely used in fields such as wildlife conservation, wild sheep monitoring, and exploration of unknown wild environments due to their advantages such as free hovering, vertical takeoff and landing, and flexible control. However, the following three problems exist in their application:

[0022] 1. The decision-making ability of intelligent flying cars in complex working environments is crucial. Among them, the ability to complete path planning and efficiently execute pre-set tasks under existing information is a critical technical aspect. Currently, most intelligent flying cars are controlled by operators, resulting in a low level of automation in autonomous path planning at the practical application level. Some amphibious platforms perform single path planning based on the controller's settings before mission execution. This planning is not applicable to some complex application scenarios and makes it difficult to guarantee the maneuverability and energy efficiency of the flying car. Furthermore, limited by mode switching, it cannot effectively adjust autonomously according to terrain, making it difficult to achieve agile, flexible, and real-time adjustments to path planning strategies.

[0023] 2. During field operations, intelligent flying cars often become uncontrollable due to GPS signal loss, remote control signal loss, or other issues. In such situations, finding an area where the flying car can land safely and autonomously becomes crucial. Flying cars can reach a distance of about five kilometers from the user, but in this range, interference from the outdoor environment often leads to GPS and remote control signal loss. Since flying cars lack the necessary emergency fully automatic landing capabilities, this can result in damage or loss to the user's vehicle. Currently, using radar for autonomous landing is effective, but radar is expensive and heavy, making it unsuitable for deployment on small to medium-sized flying cars. Overall, there is currently a lack of cost-effective, accurate, stable, and efficient autonomous landing methods for flying cars.

[0024] 3. Intelligent flying car control systems are mostly controlled by computer ground stations or remote controllers, using traditional communication direct-drive control methods. They cannot form a control system that can autonomously judge outputs and conduct risk assessments of control, and have problems with low safety and reliability.

[0025] To address the aforementioned technical issues, this application provides an intelligent flying car with a path planning system. By setting a starting point and target point in a DEM (Depth Elevation Model), and marking dangerous areas such as cloud formations, bird flocks, and heavy rain, as well as the area where the mission needs to be performed, the system uses 2D-A* and 3D-A* algorithms to determine the mission route for ground driving and aerial flight within the global map model. When a mode switch is required, the BAS (Body Assignment System) algorithm is used to optimize the switching point within a certain area around it. Upon reaching the vicinity of the mode switch point, the platform's onboard sensing equipment accurately maps the local area, identifies potential hazards, and then uses the BAS for real-time decision-making to complete the overall mission route. This avoids the poor real-time performance and limited number of mode switch points inherent in 2D-A* and 3D-A* planning methods.

[0026] Equipped with an autonomous landing system, when the intelligent flying car approaches the target point and loses its positioning signal, it fuses RGB and depth images to enhance perception. The landing point detection algorithm, which combines depth and RGB images, compensates for both the insufficient depth of RGB images and the lack of visual information in depth images. By fusing discrete information, steepness information, color information, and the Euclidean distance from the center of the optimal safe landing area to the flying car's location, a score for the safe landing point is obtained. All safe landing points are sorted according to their scores to obtain the optimal safe landing point, thus deriving the trajectory planning for the flying car. A novel method for calculating the safe landing point score is proposed: discreteness is used to detect whether the area is stable, steepness is used to detect whether the area is uneven, Euclidean distance is used to detect the relative distance between the area and the flying car, and color is used to detect whether the area is ground. These are fused to obtain the safe area score, and finally, the center point of the safe area is set as the safe landing point.

[0027] It features a redundant control system that combines external control information with sensor data, effectively solving the problem of misoperation under human intervention and providing higher safety assurance for the operation of flying cars; it also addresses the lack of existing technologies for providing multiple control inputs to flying cars and risk assessment of control information.

[0028] The advantages of the intelligent flying car proposed in this application are:

[0029] 1. Improve the computation speed in 3D path planning scenarios, and select different cost functions based on the characteristics of the task to achieve intelligent path planning between the starting point and the target point, ensuring the targeted completion of the existing task; the modal switching point decision method includes search and landing point gradient optimization based on the beetle whisker algorithm, further setting the search space and the safe distance from obstacles such as extreme weather or bird flocks, to further improve and optimize the actual applicability of the switching point based on the modal switching point determined by the existing A* algorithm;

[0030] 2. The intelligent flying car of this application can be applied in complex terrain areas such as high mountains, Gobi deserts, outposts, and high-altitude and cold regions, involving unmanned material transportation, rescue and reconnaissance in special terrain scenarios, etc.; it has high intelligence and flexibility in the face of complex field application scenarios, reliability and mobility in the process of mission execution, improves energy utilization efficiency, and protects comprehensive performance indicators such as battery SOC and SOH.

[0031] 3. Experimental data show that, under the same starting point and destination, the path planning system of this application has less time and energy consumption in overall route planning, and can plan safer and more flexible routes.

[0032] 4. The autonomous landing system of this application fuses RGB images with depth images to enhance perception capabilities. The landing point detection algorithm that fuses depth and RGB images can compensate for both the lack of depth in RGB images and the lack of visual information in depth images. By fusing the discrete information, steepness information, color information, and Euclidean distance from the center of the safe landing area to the location of the flying car, a score for the safe landing point is obtained. All safe landing points are sorted according to their scores to obtain the optimal safe landing point, thereby deriving the trajectory planning of the flying car. A new method for calculating the score of the safe landing point is proposed. Discreteness is used to detect whether the area is stable, steepness is used to detect whether the area is uneven, Euclidean distance is used to detect the relative distance between the area and the flying car, and color is used to detect whether the area is the ground. These are fused to obtain the safe area score, and finally, the center point of the safe area is set as the safe landing point.

[0033] 5. The intelligent flying car of this application combines external control information with sensor equipment information, effectively solving the problem of misoperation under human intervention control and providing higher safety assurance for the operation of the flying car.

[0034] After introducing the application scenarios and design concepts of the embodiments of this application, the technical solutions provided by the embodiments of this application will be described below.

[0035] like Figure 1 As shown in the illustration, this application provides an intelligent flying car, which is equipped with a positioning terminal and two monocular RGB cameras or one monocular RGB camera and one depth camera. The positioning terminal is a GPS terminal or a BeiDou terminal. The intelligent flying car includes:

[0036] The path planning system is used to plan the path for ground driving and air flight using elevation information during the movement of an intelligent flying car from the starting point to the target point. At the same time, it calculates the optimal mode switching point and controls the mode switching from ground driving to air flight or from air flight to ground driving.

[0037] An autonomous landing system is used to obtain the optimal safe landing point by using simultaneous RGB and depth images in space and time when the intelligent flying car approaches the target point and cannot receive a positioning signal, and then uses the final safe landing point to control the flying car to land safely.

[0038] In this embodiment, the path planning principle is as follows: ground travel is the primary method, aiming to travel the longest possible distance on the ground to conserve energy. Aerial travel is secondary, employing flight mode to overcome the limitations of harsh terrain when ground travel is impossible.

[0039] like Figure 2 As shown, the path planning system includes:

[0040] The acquisition unit is used to acquire the current mode and actual position of the intelligent flying car;

[0041] In this embodiment, the intelligent flying car is a platform that primarily operates on the ground and is supplemented by aerial flight. The modes of the intelligent flying car include ground driving and aerial flight. At the starting point, the intelligent flying car is in the mode of ground driving.

[0042] First judgment unit: used to determine whether the current mode of the intelligent flying car is ground driving. If it is, it enters the ground search unit; otherwise, it enters the second judgment unit.

[0043] Ground search unit: Based on the actual position of the intelligent flying car at the current moment, it uses a preset 2D-A* algorithm to search for the ground reachable position of the intelligent flying car at the next moment, and determines whether it can be found. If it can, the ground reachable position of the intelligent flying car at the next moment is sent to the ground driving controller; otherwise, it enters the first mode switching unit.

[0044] A* is a heuristic search algorithm with high search efficiency, especially in handling static scenarios where it demonstrates excellent path planning capabilities. The 2D-A* algorithm: When the vehicle is performing a mission in the field and driving on the ground, the planner of the intelligent flying car performs a latitude and longitude search in a plane based on a cost function. Compared to traditional algorithms in urban or laboratory environments, the 2D-A* algorithm in this embodiment matches the latitude and longitude of a node with its corresponding altitude information during the node search process. That is, it uses a two-dimensional search framework to solve three-dimensional scenarios with significant altitude fluctuations in the field, improving the algorithm's computational efficiency. In addition to the cost function, the 2D-A* algorithm also includes node feasibility analysis, including the maximum turning angle gradient, maximum tilt gradient, and maximum pitch gradient that the platform can handle. These parameters are derived from actual mobile platform maneuverability experiments.

[0045] Specifically, based on the current location of the intelligent flying car, a pre-defined 2D-A* algorithm is used to search for the ground-reachable location of the intelligent flying car at the next moment; including:

[0046] Obtain the grid containing the actual location of the intelligent flying car at the current moment from the rasterized map;

[0047] Obtain the latitude and longitude of the center points of the eight adjacent grids around the grid, and use the center points of the adjacent grids as candidate nodes;

[0048] Acquire the elevation data of all candidate points, calculate the turning angle gradient, tilt gradient and pitch angle gradient between the actual location point and each candidate point, and determine whether the actual location point is reachable from each candidate node. If all candidate nodes are unreachable, the ground reachable location of the intelligent flying car at the next moment cannot be found.

[0049] Otherwise, calculate the cost function between the actual location point and the reachable candidate node, and take the reachable candidate node with the smallest cost function as the ground reachable location of the intelligent flying car in the next moment.

[0050] The process of determining whether the actual location point is reachable from each candidate node includes:

[0051] Determine whether the following expression is true:

[0052]

[0053] Wherein, the turning angle gradient between the actual location point and the candidate point is gx, the tilt gradient is gy, and the pitch angle gradient is gz; gx min and gx max For the minimum and maximum values ​​of the turning angle gradient; gy min and gy max The minimum and maximum values ​​of the gradient of inclination; gz min and gz max These represent the minimum and maximum values ​​of the pitch angle gradient.

[0054] This includes calculating the cost function between the actual location point and reachable candidate nodes; including:

[0055] Calculate the actual cost function value G(n) between the actual location point and the nth reachable candidate node:

[0056] G(n) = D(n) + E(n)

[0057] Where D(n) is the distance between the actual location and the nth reachable candidate node, and E(n) is the driving energy consumption:

[0058]

[0059] Where μ is the coefficient of friction, m is the mass of the intelligent flying car, g is the acceleration due to gravity; Δh is the elevation change between the actual position and the nth reachable alternative node; ρ is the air density, C d Δ is the air resistance coefficient; A1 is the frontal area of ​​the vehicle on the ground; v1 is the ground speed; Δd is the horizontal distance traveled between the actual position and the nth reachable alternative node.

[0060] Calculate the estimated cost function value H(n) between the actual location point and the nth reachable candidate node:

[0061] H(n) = |x goal -x n |+|y goal -y n |+|z goal -z n |

[0062] In the formula, (x goal ,y goal ,z goal (x) represents the three-dimensional coordinates of the target point; n ,y n ,z n () represents the three-dimensional coordinates of the nth reachable candidate node;

[0063] The cost function F(n) between the actual location point and the reachable candidate node is:

[0064] F(n) = G(n) + H(n).

[0065] First mode switching unit: used to obtain the optimal mode switching point for switching from ground driving to air flight using a preset BAS algorithm, and send the optimal mode switching point to the mode switcher of the intelligent flying car.

[0066] like Figure 3 As shown, the BAS algorithm is a heuristic algorithm with strong robustness. It can achieve efficient optimization without prior processing or environmental obstacle modeling and analysis, even without knowing the specific form of the function, making it highly suitable for the practical application scenario of this embodiment. This embodiment can avoid dangerous situations such as rollover and shutdown caused by the A* algorithm determining unfavorable mode switching point conditions for platform takeoff and landing by optimizing the mode switching point.

[0067] The optimal mode switching point for transitioning from ground driving to air flight is obtained using a pre-defined BAS algorithm, including:

[0068] Step S1: Generate random vectors

[0069]

[0070] In the formula, l is the spatial dimension, l = 3, and the three dimensions include: longitude, latitude and elevation; Rands(l,1) means to randomly generate an l-dimensional vector, and the value range of each component is [0, 1].

[0071] Step S2: Calculate the positions of the left and right whiskers of the intelligent flying car:

[0072]

[0073] In the formula, X R For the three-dimensional position of the right side of the intelligent flying car, X L For the three-dimensional position of the left whisker of the intelligent flying car, X m-1 R represents the three-dimensional position of the mode switching point in the (m-1)th iteration, where the initial mode switching point is the actual position at the current moment; R is the distance between the left and right whiskers of the intelligent flying car.

[0074] Step S3: Calculate the fitness function F(X) of the right whisker of the intelligent flying car. R ):

[0075] F(X R ) = E R +αR(X R )

[0076] Among them, E R The energy consumption of the intelligent flying car; α is a hyperparameter, and R(·) is the gradient function:

[0077] R(X R )=μ1·gx R +μ2·gy R +μ3·gz R

[0078] In the formula, gx R gy R and gz R These represent the turning angle gradient, tilt gradient, and pitch gradient of the right side of the intelligent flying car, respectively; μ1, μ2, and μ3 are all weighting factors.

[0079] Step S4: Calculate the fitness function F(X) of the left whisker of the intelligent flying car. L ):

[0080] F(X L ) = E L +αR(X L )

[0081] R(X L )=μ1·gx L +μ2·gyL +μ3·gz L

[0082] In the formula, E L For the energy consumption of the left side of the intelligent flying car, gx L gy L and gz L These are the turning angle gradient, tilt gradient, and pitch gradient of the left side of the intelligent flying car, respectively.

[0083] Step S5: Calculate the three-dimensional position X of the m-th mode switching point. m :

[0084]

[0085] In the formula, δ is the Euclidean distance of the step, and sign is the sign function;

[0086] Step S6: Determine if m is not greater than the preset iteration number M. If yes, increment m by 1 and proceed to step S1; otherwise, set X... M As the kth mode switching point, the initial value of k is 1, proceed to step S7;

[0087] Step S7: Determine whether k is not greater than the preset number of searches K. If yes, increment k by 1 and proceed to step S1; otherwise, proceed to step S8.

[0088] Step S8: Calculate the fitness function of the K mode switching points, and take the mode switching point with the smallest fitness function as the optimal mode switching point.

[0089] The second judgment unit is used to obtain the battery SOC value and total flight distance of the intelligent flying car during flight, and to determine whether either the battery SOC value is greater than a preset power percentage or the ratio of total flight distance to total mission distance is greater than a preset distance percentage. If yes, it enters the second mode switching unit; otherwise, it enters the air search unit.

[0090] Obtain the battery SOC value of the intelligent flying car during flight, including:

[0091] The SOC (t) value of the battery of the intelligent flying car during flight is:

[0092]

[0093] In the formula, Q0 is the initial battery energy, Q is the nominal battery energy, T is the sequence number of the current sampling time, and E t The energy consumption from the (t-1)th sampling time to the tth sampling time;

[0094] Where, when the journey from the (t-1)th sampling time to the tth sampling time is ground driving, then E t Energy consumption during driving;

[0095] When the transition from ground driving to aerial flight occurs between the (t-1)th sampling time and the tth sampling time, the energy consumption E t for:

[0096] E t =E expand +E Bodeneffekt

[0097] E expand The energy consumed for an intelligent flying car to deploy its wings; E Bodeneffekt Energy consumed by the ground effect;

[0098] If the flight occurs from the (t-1)th sampling time to the tth sampling time, then E t For flight energy consumption;

[0099] When the transition from the (t-1)th sampling time to the tth sampling time is a mode switch from airborne flight to ground travel, the energy consumption E t for:

[0100] E t =E fold +E Bodeneffekt

[0101] E fold The energy consumed by folding the wings of an intelligent flying car.

[0102] Preferably, the preset battery percentage is 40% and the preset distance percentage is 50%.

[0103] The second mode switching unit uses a preset BAS algorithm to obtain the optimal mode switching point from air flight to ground driving, and sends the optimal mode switching point to the mode switcher of the intelligent flying car.

[0104] The process of obtaining the optimal mode switching point from air flight to ground driving using the preset BAS algorithm is the same as the process of obtaining the optimal mode switching point from ground driving to air flight.

[0105] If the mode switching point is too close to the target point, it will automatically fly directly to the target point according to the established mode switching rules, so as to reduce the energy consumption caused by frequent mode switching and calculation, and complete the field rescue and reconnaissance mission without damaging the health and life of the components.

[0106] Airborne search unit: Based on the current actual location of the intelligent flying car, it uses a preset 3D-A* algorithm to search for the airborne reachable location of the intelligent flying car in the next moment; and sends the airborne reachable location of the intelligent flying car in the next moment to the airborne flight controller.

[0107] In this embodiment, based on the current actual position of the intelligent flying car, the 3D-A* algorithm is used to search for the next available aerial position of the intelligent flying car; including:

[0108] Obtain the grid containing the actual location of the intelligent flying car at the current moment from the rasterized map;

[0109] Obtain the latitude, longitude, and elevation of the center points of the eight adjacent grids around the grid, and use the center points of the eight adjacent grids as eight candidate nodes;

[0110] Calculate the cost function between the actual location point and the eight candidate nodes, and obtain the candidate node with the smallest cost function;

[0111] The elevation of the candidate node with the minimum cost function is corrected by using a pre-set virtual altitude variable Δz, and the position of the corrected candidate node is used as the air reachable position of the intelligent flying car in the next moment.

[0112] The calculation of the cost function between the actual location point and the eight candidate nodes includes:

[0113] Calculate the actual cost function value G(n) between the actual location point and the nth candidate node:

[0114] G(n) = D(n) + E(n)

[0115] Where D(n) is the distance between the nth candidate node and the actual location, and E(n) is the flight energy consumption:

[0116] E(n)=E hover +E move

[0117] Among them, hovering energy E hover for:

[0118]

[0119] Where ρ is the air density, m is the mass of the intelligent flying car, g is the gravitational acceleration; b is the number of propellers of the intelligent flying car, r is the radius of the propeller, η is the motor efficiency; v2 is the flight speed of the intelligent flying car.

[0120] Mobile energy E move for:

[0121]

[0122] Where Δh is the elevation change between the actual location point and the nth candidate node; A2 is the flight frontal area; Δd is the horizontal movement distance between the actual location point and the nth candidate node; C d This refers to the air drag coefficient;

[0123] Calculate the estimated cost function value H(n) between the actual location point and the nth candidate node:

[0124] H(n) = |x goal -x n |+|y goal -y n |+|z goal -z n |

[0125] In the formula, (x goal ,y goal ,z goal (x) represents the three-dimensional coordinates of the target point; n ,y n ,z n () represents the three-dimensional coordinates of the nth candidate node;

[0126] The cost function F(n) between the actual location point and the nth candidate node is:

[0127] F(n) = G(n) + H(n).

[0128] The parameters of the 2D-A* and 3D-A* algorithms in this embodiment are shown in Table 1:

[0129] Table 1 Parameters in the equation

[0130]

[0131] This embodiment introduces a virtual altitude variable Δz. Specifically, 3D A* is divided into three phases: takeoff, cruise, and landing. In the takeoff phase, the mobile platform is required to fly upwards along the target direction. During this phase, the Δz value smoothly increases as the platform moves forward until it reaches the value of the virtual altitude variable Δz. In the cruise phase, the flight altitude is the altitude (z) of the nodes passed through plus the value of Δz. When the platform approaches the target node or mode switching point, it enters the landing phase, requiring the mobile platform to descend downwards along the target direction. During this phase, the Δz value smoothly decreases as the platform moves forward until it reaches the z value of the target node or mode switching point.

[0132] The aforementioned virtual height can be flexibly adjusted according to the elevation fluctuations of the terrain near the task, ensuring that the mobile platform will not touch obstacles such as mountains at this height, thus avoiding dangerous situations. The simulation results after adding Δz are as follows: Figure 4 As shown in the figure, z max and z min These are the maximum and minimum flight altitudes set for the platform, respectively, with "ground" representing the ground altitude. max This setting can limit the platform's maximum flight altitude, preventing energy waste caused by excessive flight altitude; min The settings can limit the lower limit of the platform's flight altitude to avoid dangerous situations such as the platform colliding with mountains and vegetation due to excessively low flight altitude.

[0133] Specifically, the elevation of the candidate node with the minimum cost function is corrected using a pre-set virtual height variable Δz. This includes: if the elevation of the candidate node with the minimum cost function is z, then the corrected elevation is z + Δz.

[0134] To apply the BAS algorithm to path planning based on the 3D-A* algorithm, this embodiment uses a DEM data map. The DEM map does not have traditional obstacle zones; it mainly consists of maneuverable zones, non-maneuverable zones, and recommended avoidable danger zones. The delineation of maneuverable and non-maneuverable zones is determined by factors such as the slope, gradient, and inclination of the corresponding grid. Areas that meet the platform's maneuverability requirements are considered maneuverable zones and fall within the planning scope of the optimization algorithm; otherwise, they are not. The longhorn beetle must be able to move freely within the maneuverable zone, avoid non-maneuverable zones, and try to avoid danger zones. The map values ​​corresponding to the xyz coordinate axes are calculated using the gradient function.

[0135] In addition, the method also includes: using path points and decision points as decision points of Bézier curves, performing path smoothing processing, and finally outputting the path planning results of the land-air amphibious multimodal rescue platform in the field mission environment.

[0136] Using the path planning method of this application, under the same task conditions, the method of this application has a more time-saving and labor-saving path output after the mode switching point is updated. It has very necessary practical application significance for platform rescue work in the wild environment. The remaining power can be used for the return trip or for the transmission of key information such as GPS video and pictures.

[0137] The BAS modality switching point was set to iterate 20 steps per iteration, with a step size of 1 meter. A total of 80 target points were searched. Two danger zones were established on the map. The optimized path output by the path planning system effectively avoided the danger zones, which is of great significance for safety in the actual application of the platform. Furthermore, the optimized path reduced the total time spent on the task, helping to improve the platform's environmental adaptability and maneuverability.

[0138] Table 2 Comparison of Path Planning Algorithm Mode Switching Point Calculation Time

[0139]

[0140] like Figure 5 As shown, the autonomous landing system includes:

[0141] The image acquisition unit is used to acquire simultaneous spatiotemporal RGB images and depth images;

[0142] This unit has two implementation methods:

[0143] The first implementation method involves installing two RGB cameras on the flying car to acquire two RGB images. Image correction is then performed on the two RGB images. The corrected RGB images are then processed to obtain a depth image. The specific process includes:

[0144] The pixel coordinate systems of the two RGB images are transformed to the camera coordinate system using a common intrinsic parameter matrix. The two camera coordinate systems are then rotated to obtain new camera coordinate systems. Distortion removal operations are performed on the left and right cameras respectively for the new camera coordinates. After the distortion removal operation is completed, the left and right camera coordinate systems are transformed back to the left and right image pixel coordinate systems using the intrinsic parameter matrices of the left and right cameras respectively. The pixel values ​​of the left and right source images are then used to interpolate the pixels of the new left and right images respectively.

[0145] After converting the calibrated RGB image to grayscale, directly take the value of the component with the largest value among the R, B, and G components, satisfying the following:

[0146]

[0147] Where p r p represents the pixel value of the R channel in an RGB image. g p represents the pixel value of the G channel in an RGB image. b is the pixel value of the B channel in the RGB image, and D(p) is the pixel value of the grayscale image;

[0148] The SAD algorithm is used to convert two grayscale images (left and right grayscale images) into a disparity map. First, the size of the SAD window, the starting matching position *p* in the left grayscale image, and the range of movement of the SAD window in the right grayscale image are set. In the left grayscale image, the position *p* of the pixel to be matched is determined, and this position is used as the anchor point of the SAD window. The SAD window covers the region *regionr* in the left grayscale image with *p* as the anchor point. In the right grayscale image, the starting point for matching is selected, position *q*, and this point is used as the anchor point of the SAD window. An image region *regionl* is formed in the right grayscale image with *q* as the anchor point.

[0149] Define difference = regionr - regionl. Calculate the sum of differences. Move the SAD along the row direction in the right-eye grayscale image (the number of moves is the size of the matching range), repeating the above steps and recording the difference obtained each time in the mat matrix. Find the minimum difference value d in the mat matrix; its position is the disparity between the right-eye and left-eye grayscale images. The disparity value satisfies the following formula:

[0150]

[0151] Where L(x+i,y+j) is the pixel value of the left-eye grayscale image, R(x+d+i,y+i) is the pixel value of the right-eye grayscale image; D(x,y,d) is the pixel value of the disparity map at pixel (x,y); x and y are the two-dimensional coordinates in the grayscale image; W and H are the image width and height.

[0152] Calculate the depth difference Z(x,y) based on the disparity D(x,y,d):

[0153]

[0154] Where f is the focal length and b0 is the distance between the centers of the two cameras, thus obtaining the depth image.

[0155] The second implementation method is to deploy an RGB camera and a depth camera on the flying car to acquire RGB images and depth images respectively.

[0156] The ground candidate region acquisition unit is used to process the smoothed depth image using the watershed algorithm to obtain the processed depth image and acquire candidate ground regions of the processed depth image.

[0157] To reduce the impact of discrete data on the watershed algorithm, the obtained depth image D is processed by opening and closing operations to obtain a smooth depth image D1 with obvious features. Since the flying car needs to land on a stable, solid, and safe ground, it is only necessary to process various information about the ground to measure whether the location meets the landing requirements. At the same time, since the entire image is no longer processed directly, but only the ground part of the depth image is processed, the computational efficiency is greatly improved.

[0158] The depth image D1 is processed using the watershed algorithm to obtain the depth image D2. Based on the pixel value p of pixel (i,j) in the depth image D1... i,j Calculate the pixel value q of pixel (i,j) in depth image D2. i,j :

[0159]

[0160] Where, p min p represents the minimum pixel value in the depth image D1. max This represents the maximum pixel value in the depth image D1.

[0161] When selecting multiple ground regions from the processed depth image, not every ground region meets the landing criteria. Due to the influence of the rotorcraft flying car's size, the flying car needs to be landed in an area that allows it to land directly on the ground. This imposes a limitation on the size of the area. Therefore, the depth image D2 needs to be processed to obtain a ground region that meets the size requirements for the flying car to land.

[0162] To this end, multiple ground regions were cropped from the depth image D2 using library functions from the OpenCV library, where the pixel value of the ground region on the depth image is 255; the largest inscribed circle of the ground region whose area is larger than that of the flying car was selected as the candidate ground region.

[0163] The calculation unit is used to calculate the mean depth variance, mean steepness, Euclidean distance, and mean color of each candidate ground region, thereby calculating the score of each candidate ground region;

[0164]

[0165]

[0166]

[0167] Where, p averageLet be the mean pixel value of the candidate ground region in depth image D1, I be the number of row pixels in the candidate ground region in depth image D1, J be the number of column pixels in the candidate ground region in depth image D1, i be the row pixel, and j be the column pixel; Let D1 be the normal to the 3D surface of the depth image. The z-axis vector of the three-dimensional surface;

[0168] Calculate the distance from the center point of the candidate ground region in depth image D to the center point p of depth image D. x =(x max / 2,y max / 2) Euclidean distance S euclidean ;where x max and y max This is the maximum pixel size of the depth map; for example, 640x480, where x... max and y max They are 640 and 480 respectively.

[0169] Because in the wild, relatively safe ground is generally bare mud, the color score S for each candidate ground area is... color Satisfy the following formula:

[0170]

[0171] Where C color G is the average color value of the candidate ground region, obtained from the RGB image (left-eye RGB image or right-eye RGB image); G is the set average ground color value.

[0172] Calculate the score S for each candidate ground region. score :

[0173] S score =λ1S steepness +λ2S depth +λ3S euclidean +λ4S color

[0174] λ1, λ2, λ3, and λ4 are weighting coefficients that satisfy:

[0175] λ1+λ2+λ3+λ4=1

[0176] Preferably, the value of λ1 is 0.3, the value of λ2 is 0.2, the value of λ3 is 0.2, and the value of λ4 is 0.3.

[0177] The optimal safe landing point acquisition unit is used to sort the center points of all candidate ground areas in descending order of score, and select the center point with the highest score as the optimal safe landing point.

[0178] The control unit is used to determine whether the distance between the optimal safe landing point and the center point of the RGB image is less than a preset threshold. If it is, the optimal safe landing point is used as the final safe landing point, and the flying car is controlled to land safely using the final safe landing point. Otherwise, the flying car is controlled to move towards the optimal safe landing point, and the optimal safe landing point is searched again during the movement.

[0179] In addition, controlling the flying car to move toward the optimal safe landing point also includes: using multiple infrared sensors to detect and avoid obstacles around the flying car.

[0180] As one possible implementation, the intelligent flying car in this embodiment further includes: a redundant control system, used to determine whether there is a risk in the external control information and generate the final control information when receiving at least one of the external control information, such as remote controller information, joystick information and ground station control information.

[0181] like Figure 6 As shown, the redundant control system includes: a remote control information acquisition unit, a joystick information acquisition unit, a ground station control information acquisition unit, sensor devices, and a redundant control decision unit;

[0182] The remote control information acquisition unit is used to receive control command information issued by the remote control and generate first control information. Specifically, the remote control information acquisition unit includes a receiver and a decoding unit. The receiver is used to receive the control command information of the remote control via SBUS protocol radio signals. The decoding unit is used to decode the control command information to obtain the first control information and encode the first control information into a PPM protocol electrical signal to send it to the redundant control decision unit.

[0183] The joystick information acquisition unit is used to receive commands generated by joystick movement and perform line recognition to obtain second control information; specifically, the joystick information acquisition unit is used to: receive control command information of the joystick, recognize the control command information of the joystick to obtain second control information, and encode the second control information into a PPM protocol electrical signal and send it to the redundant control decision unit.

[0184] The ground station control information acquisition unit is used to receive third control information sent by the ground station; wherein, the ground control station is equipped with a control unit, a mission planning unit, and a data transmission unit; the control unit is used for operators to set control command information and send it to the mission planning unit; the mission planning unit is used to integrate the control command information to generate third control information and send it to the data transmission unit; the data transmission unit is used to send the third control information to the ground station control information acquisition unit via a wireless link.

[0185] The sensor device is used to collect the sensing information of the flying car; the sensor device includes: an inertial measurement unit, a magnetic compass, a barometer, a vision sensor, and an airspeed barometer.

[0186] The redundant control decision unit is used to obtain the final control information according to a preset strategy based on the sensing information and the received control information of the flying car, and to generate control action commands for the flying car based on the final control information; the control action commands are sent to the action execution unit of the flying car via the CAN bus; the sensing information includes: the position information, attitude information, speed information, geomagnetic information and air pressure information of the flying car.

[0187] The redundant control decision unit is specifically used for:

[0188] Acquire control information, wherein the control information includes at least one of first control information, second control information, and third control information;

[0189] Using each control information, a simulated action is generated. Based on the current position, attitude, and speed information of the flying car and the simulated action, the position, attitude, and speed information of the flying car at the next moment are predicted. Preferably, the difference between the next moment and the current moment is 0.5 seconds.

[0190] Based on the position, attitude, and speed information of the flying car at the next moment corresponding to each control information, a risk assessment condition is used to determine whether the control information poses a risk. The risk assessment condition includes at least one of the following: the speed of the flying car at the next moment is greater than a preset speed threshold, the pitch angle of the flying car at the next moment is greater than a preset pitch angle threshold, the roll angle of the flying car at the next moment is greater than a preset roll angle threshold, and the distance between the position of the flying car and the obstacle at the next moment is less than a preset safe distance.

[0191] Based on whether each control information poses a risk and the preset priority of the control information, the final control information is obtained, and the control action commands for the flying car are generated using the final control information.

[0192] Specifically, based on whether each control information poses a risk and the preset priority of the control information, the final control information is obtained, including:

[0193] If all control information poses a risk, the control information from the previous moment will be used as the final control information; that is, the flying car will maintain its original flight maneuvers.

[0194] If only one piece of control information poses no risk, then that control information shall be used as the final control information.

[0195] If there are two or three control information items that do not pose a risk, the unique control information obtained from the two or three control information items shall be used as the final control information according to the priority of the first control information, the third control information, and the second control information.

[0196] The redundant control system further includes a data transmission unit for sending feedback data to the data transmission unit; the feedback data includes at least one of the following: the flying car's position information, attitude information, speed information, control action command information, and system operation information; the ground control station further includes a display unit for visualizing and displaying the feedback data. Through information visualization, control parameters and various parameters of the flying car can be better monitored, greatly improving the safety and reliability of the flying car.

[0197] The flying car's motion execution unit is used to control the flying car according to its control commands. This unit includes a flight motion execution subunit and a driving motion execution subunit.

[0198] The flight action execution subunit is used to receive control action commands and execute corresponding aerial flight actions, including: takeoff, landing, hovering, forward movement, backward movement, left and right translation, and rotation; the flight action execution unit includes a propeller motor;

[0199] The driving action execution subunit is used to receive control action commands and execute corresponding ground driving actions, including: starting, moving forward, reversing, and turning. The driving action execution unit includes: wheel motors, steering motors, and brake motors.

[0200] It should be noted that although several units or sub-units of the device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units.

[0201] Furthermore, although the operations of the method of this application are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0202] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit it. Although this application has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent substitutions to the technical solutions of this application do not depart from the spirit and scope of the technical solutions of this application, and should all be covered within the scope of the claims of this application.

Claims

1. An intelligent flying car, wherein the intelligent flying car is equipped with a positioning terminal and two monocular RGB cameras or one monocular RGB camera and one depth camera, characterized in that, The intelligent flying car includes: The path planning system is used to plan the path for ground driving and air flight using elevation information during the movement of an intelligent flying car from the starting point to the target point. At the same time, it calculates the optimal mode switching point and controls the mode switching from ground driving to air flight or from air flight to ground driving. An autonomous landing system is used to obtain the optimal safe landing point by using RGB and depth images of the same space and time when the intelligent flying car approaches the target point and cannot receive a positioning signal, and to control the flying car to land safely using the final safe landing point. The path planning system includes: The acquisition unit is used to acquire the current mode and actual position of the intelligent flying car; First judgment unit: used to determine whether the current mode of the intelligent flying car is ground driving. If it is, it enters the ground search unit; otherwise, it enters the second judgment unit. Ground search unit: Based on the actual position of the intelligent flying car at the current moment, it uses a preset 2D-A* algorithm to search for the ground reachable position of the intelligent flying car at the next moment, and determines whether it can be found. If it can, the ground reachable position of the intelligent flying car at the next moment is sent to the ground driving controller; otherwise, it enters the first mode switching unit. First mode switching unit: used to obtain the optimal mode switching point for switching from ground driving to air flight using a preset BAS algorithm, and send the optimal mode switching point to the mode switcher of the intelligent flying car. The second judgment unit is used to obtain the battery SOC value and total flight distance of the intelligent flying car during flight, and to determine whether either the battery SOC value is greater than a preset power percentage or the ratio of total flight distance to total mission distance is greater than a preset distance percentage. If yes, it enters the second mode switching unit; otherwise, it enters the air search unit. The second mode switching unit is used to obtain the optimal mode switching point from air flight to ground driving using a preset BAS algorithm, and send the optimal mode switching point to the mode switcher of the intelligent flying car. Airborne search unit: Based on the current actual location of the intelligent flying car, it uses a preset 3D-A* algorithm to search for the airborne reachable location of the intelligent flying car in the next moment; and sends the airborne reachable location of the intelligent flying car in the next moment to the airborne flight controller. The optimal mode switching point for transitioning from ground driving to air flight is obtained using a pre-defined BAS algorithm, including: Step S1: Generate random vectors : In the formula, For spatial dimensions, The three dimensions include: longitude, latitude, and elevation; Indicates random generation A dimensional vector, where each component takes values ​​in the range [0, 1]. Step S2: Calculate the positions of the left and right whiskers of the intelligent flying car: In the formula, The three-dimensional position of the right side of the intelligent flying car. The three-dimensional position of the left whisker of the intelligent flying car. Indicates the first The three-dimensional position of the mode switching point in the next iteration, where the initial mode switching point is the actual position point at the current moment; The distance between the left and right whiskers of the intelligent flying car; Step S3: Calculate the fitness function of the right whisker of the intelligent flying car. : in, The energy consumption required for intelligent flying cars; For hyperparameters, Gradient function: In the formula, , These are the turning angle gradient, tilt gradient, and pitch gradient of the right side of the intelligent flying car, respectively. , and All are weighting factors; Step S4: Calculate the fitness function of the left whisker of the intelligent flying car. : In the formula, The energy consumption required for intelligent flying cars , These are the turning angle gradient, tilt gradient, and pitch gradient of the left side of the intelligent flying car, respectively. Step S5: Calculate the three-dimensional position of the m-th mode switching point. : In the formula, For the step Euclidean distance, ... It is a symbolic function; Step S6: Determine if m is not greater than the preset iteration number M. If yes, increment m by 1 and proceed to step S1; otherwise, ... As the kth mode switching point, the initial value of k is 1, proceed to step S7; Step S7: Determine whether k is not greater than the preset number of searches K. If yes, increment k by 1 and proceed to step S1; otherwise, proceed to step S8. Step S8: Calculate the fitness function of the K mode switching points, and take the mode switching point with the smallest fitness function as the optimal mode switching point.

2. The intelligent flying car according to claim 1, characterized in that, Based on the current location of the intelligent flying car, a pre-defined 2D-A* algorithm is used to search for the ground-reachable location of the intelligent flying car at the next moment; including: Obtain the grid containing the actual location of the intelligent flying car at the current moment from the rasterized map; Obtain the latitude and longitude of the center points of the eight adjacent grids around the grid, and use the center points of the adjacent grids as candidate nodes; Acquire the elevation data of all candidate points, calculate the turning angle gradient, tilt gradient and pitch angle gradient between the actual location point and each candidate point, and determine whether the actual location point is reachable from each candidate node. If all candidate nodes are unreachable, the ground reachable location of the intelligent flying car at the next moment cannot be found. Otherwise, calculate the cost function between the actual location point and the reachable candidate node, and take the reachable candidate node with the smallest cost function as the ground reachable location of the intelligent flying car in the next moment.

3. The intelligent flying car according to claim 2, characterized in that, Determine whether the actual location point is reachable from each candidate node, including: Determine whether the following expression is true: Wherein, the turning angle gradient between the actual location point and the candidate point is The gradient of the inclination is The pitch angle gradient is ; and These represent the minimum and maximum values ​​of the turning angle gradient; and These represent the minimum and maximum values ​​of the gradient of inclination. and These represent the minimum and maximum values ​​of the pitch angle gradient.

4. The intelligent flying car according to claim 3, characterized in that, Calculate the cost function between the actual location point and reachable candidate nodes; including: Calculate the actual cost function value between the actual location point and the nth reachable candidate node. : in, This represents the distance between the actual location point and the nth reachable candidate node. Energy consumption for driving: in, The coefficient of friction, Let g be the mass of the intelligent flying car, and g be the acceleration due to gravity. This represents the elevation change between the actual location point and the nth reachable candidate node. air density, This refers to the air drag coefficient; The frontal area for ground driving; Ground speed; This represents the horizontal distance between the actual location and the nth reachable alternative node. Calculate the estimated cost function value between the actual location point and the nth reachable candidate node. : In the formula, ( These are the three-dimensional coordinates of the target point; () represents the three-dimensional coordinates of the nth reachable candidate node; The cost function between the actual location point and the reachable candidate node. for: 。 5. The intelligent flying car according to claim 1, characterized in that, Based on the current location of the intelligent flying car, the 3D-A* algorithm is used to search for the next available location in the air for the intelligent flying car; including: Obtain the grid containing the actual location of the intelligent flying car at the current moment from the rasterized map; Obtain the latitude, longitude, and elevation of the center points of the eight adjacent grids around the grid, and use the center points of the eight adjacent grids as eight candidate nodes; Calculate the cost function between the actual location point and the eight candidate nodes, and obtain the candidate node with the smallest cost function; Using pre-set virtual height variables z corrects the elevation of the candidate node with the minimum cost function and uses the corrected position of the candidate node as the reachable position of the intelligent flying car in the next moment.

6. The intelligent flying car according to claim 5, characterized in that, Calculate the cost function between the actual location point and the eight candidate nodes, including: Calculate the actual cost function value between the actual location point and the nth candidate node. : in, Let n be the distance between the nth candidate node and the actual location. Energy consumption for flight: Among them, hovering energy for: in, air density, Let g be the mass of the intelligent flying car, and g be the acceleration due to gravity. Let r be the number of propellers in the intelligent flying car, and r be the radius of the propeller. For motor efficiency; The flight speed of the intelligent flying car; Mobile energy for: in, This represents the elevation change between the actual location point and the nth candidate node. For the windward area of ​​the flight; This represents the horizontal distance between the actual location and the nth candidate node. This refers to the air drag coefficient; Calculate the estimated cost function value between the actual location point and the nth candidate node. : In the formula, ( These are the three-dimensional coordinates of the target point; () represents the three-dimensional coordinates of the nth candidate node; The cost function between the actual location point and the nth candidate node. for: 。 7. The intelligent flying car according to claim 6, characterized in that, Obtain the battery SOC value of the intelligent flying car during flight, including: Battery SOC value of intelligent flying cars during flight for: In the formula, Initial battery energy, The nominal battery energy, This is the sequence number of the current sampling time. For the first From the sampling time to the... Energy consumption at each sampling time; Among them, when the first From the sampling time to the... If the sampling time is during ground driving, then Energy consumption during driving; When the From the sampling time to the... If each sampling moment represents the mode transition from ground travel to aerial flight, then the energy consumption... for: The energy consumed to deploy the wings of an intelligent flying car; Energy consumed by the ground effect; When the From the sampling time to the... If the sampling time is during flight, then For flight energy consumption; When the From the sampling time to the... If each sampling moment represents the mode transition from airborne flight to ground-based travel, then the energy consumption... for: The energy consumed by folding the wings of an intelligent flying car.

8. The intelligent flying car according to claim 1, characterized in that, The autonomous landing system includes: The image acquisition unit is used to acquire simultaneous spatiotemporal RGB images and depth images; The ground candidate region acquisition unit is used to process the smoothed depth image using the watershed algorithm to obtain the processed depth image and acquire candidate ground regions of the processed depth image. The calculation unit is used to calculate the mean depth variance, mean steepness, Euclidean distance, and mean color of each candidate ground region, thereby calculating the score of each candidate ground region; The optimal safe landing point acquisition unit is used to sort the center points of all candidate ground areas in descending order of score, and select the center point with the highest score as the optimal safe landing point. The control unit is used to determine whether the distance between the optimal safe landing point and the center point of the RGB image is less than a preset threshold. If it is, the optimal safe landing point is used as the final safe landing point, and the flying car is controlled to land safely using the final safe landing point. Otherwise, the flying car is controlled to move towards the optimal safe landing point, and the optimal safe landing point is searched again during the movement.

9. The intelligent flying car according to claim 8, characterized in that, The ground candidate region acquisition unit is specifically used for: For depth images Opening and closing operations are performed to obtain a smooth depth image with distinct features. ; Based on depth image pixels Calculate depth image pixels : in, Representing depth image The minimum pixel value in Representing depth image The maximum pixel value in; From depth image Multiple ground regions are obtained, where each ground region has a pixel value of 255 in the depth image; The number of rows in pixels. The number of rows in pixels; The largest inscribed circle of the ground region whose area is larger than that of the flying car is selected as the candidate ground region.

10. The intelligent flying car according to claim 9, characterized in that, The computing unit is specifically used for: For depth images For each candidate ground region, calculate the mean depth variance. and the mean steepness : in, For depth images The mean value of the pixel values ​​in the candidate ground region. For depth images The number of row pixels in the candidate ground region. For depth images The number of column pixels in the candidate ground region. For row pixels, For column pixels; For depth images Normals of a three-dimensional surface For three-dimensional curved surfaces Axis vectors; Calculate depth image Center point of candidate ground region to depth image Euclidean distance of the center point ; Color score for each candidate ground region for: in, The color mean of the candidate ground region is obtained from the RGB image. The set average ground color; The score of the candidate ground region for: , , and Let the weighting coefficients satisfy: 。 11. The intelligent flying car according to claim 1, characterized in that, The intelligent flying car also includes a redundant control system, which, when receiving at least one external control information from remote controller information, joystick information, and ground station control information, determines whether the external control information poses a risk and generates final control information.

12. The intelligent flying car according to claim 11, characterized in that, The redundant control system includes: a remote control information acquisition unit, a joystick information acquisition unit, a ground station control information acquisition unit, sensor devices, and a redundant control decision unit; The remote control information acquisition unit is used to receive control command information from the remote control via SBUS protocol radio signals, decode the control command information to obtain first control information, and encode the first control information into PPM protocol electrical signals and send them to the redundant control decision unit. The joystick information acquisition unit is used to receive control command information from the joystick, identify the control command information from the joystick to obtain second control information, and encode the second control information into a PPM protocol electrical signal and send it to the redundant control decision unit. The ground station control information acquisition unit is used to receive third control information sent by the ground station; The sensor equipment includes: an inertial measurement unit, a magnetic compass, a barometer, a vision sensor, and an airspeed / barometer; The redundant control decision unit is used to obtain the final control information according to a preset strategy based on the sensing information of the flying car and the received control information, generate control action commands for the flying car according to the final control information, and send the control action commands to the action execution unit of the flying car. The sensing information includes: the position information, attitude information, speed information, geomagnetic information and air pressure information of the flying car.

13. The intelligent flying car according to claim 12, characterized in that, The ground control station is equipped with a control unit, a mission planning unit, and a data transmission unit; The control unit is used for operators to set control command information and send it to the task planning unit; The task planning unit is used to integrate control command information to generate third control information and send it to the data transmission unit. The data transmission unit is used to transmit the third control information to the ground station control information acquisition unit via a wireless link.

14. The intelligent flying car according to claim 13, characterized in that, The redundant control decision unit is specifically used for: Acquire control information, wherein the control information includes at least one of first control information, second control information, and third control information; Using each control information, simulated actions are generated. Based on the current position, attitude, and speed information of the flying car and the simulated actions, the position, attitude, and speed information of the flying car at the next moment are predicted. Based on the position, attitude, and speed information of the flying car at the next moment corresponding to each control information, a risk assessment condition is used to determine whether the control information poses a risk. The risk assessment condition includes at least one of the following: the speed of the flying car at the next moment is greater than a preset speed threshold, the pitch angle of the flying car at the next moment is greater than a preset pitch angle threshold, the roll angle of the flying car at the next moment is greater than a preset roll angle threshold, and the distance between the position of the flying car and the obstacle at the next moment is less than a preset safe distance. If all control information poses a risk, then the control information from the previous moment will be used as the final control information. If only one piece of control information is free from risk, then that control information shall be used as the final control information; if two or three pieces of control information are free from risk, then according to the priority of the first, third and second control information, the only control information obtained from the two or three pieces of control information shall be used as the final control information. The final control information is used to generate control commands for the flying car.

Images