WO2021087702A1 - Sloped terrain prediction method and device, radar, unmanned aerial vehicle, and operation control method - Google Patents

Abstract

A sloped terrain prediction method and device, a radar, an unmanned aerial vehicle, and an operation control method. The unmanned aerial vehicle is equipped with a radar. The method comprises: acquiring detection data of an omnidirectionally scanned region obtained by a process of a radar rotationally scanning the ground (S110); performing, according to the detection data, fitting to obtain a fitting plane of the omnidirectionally scanned region (S120); determining, according to the fitting plane, a terrain parameter of the omnidirectionally scanned region (S130); and adjusting, according to the terrain parameter, a maneuver of an unmanned aerial vehicle (S140).

Description

Terrain prediction method, device, radar, unmanned aerial vehicle and operation control method of slope Technical field

This specification relates to the technical field of unmanned aerial vehicles, in particular to a terrain prediction method, device, radar, unmanned aerial vehicle and operation control method for slopes.

Background technique

In some operating scenarios, drones need to fly over uneven terrain, such as sloped ground. For example, in the process of mountain operations, the UAV can detect the vertical distance to the ground directly below and maintain a certain height difference with the ground directly below, so as to realize ground-like flight such as terrain following.

However, by detecting the height difference between the drone and the ground directly below, and then adjusting the flying height, this method is a kind of post feedback, and the terrain following response is slow. Security poses a huge threat.

Summary of the invention

Based on this, this manual provides a terrain prediction method, device, radar, drone, and operation control method for slopes, which can detect all positions of the drone, such as the front, rear, left, and right ground. More comprehensively predict the terrain of the area where the drone is located.

In the first aspect, this specification provides a terrain prediction method for slopes, the method includes:

Obtain the detection data of the omnidirectional scanning area obtained by the radar during the ground rotation scanning process;

Fitting a fitting plane of the omnidirectional scanning area according to the detection data;

The terrain parameter of the omnidirectional scanning area is determined according to the fitting plane, and the terrain parameter includes the slope of the omnidirectional scanning area.

In a second aspect, the present application provides a terrain prediction device for slopes, the terrain prediction device including a memory and a processor;

The memory is used to store a computer program;

The processor is configured to execute the computer program and, when executing the computer program, implement the following steps:

Obtain the detection data of the omnidirectional scanning area obtained by the radar during the ground rotation scanning process;

Fitting a fitting plane of the omnidirectional scanning area according to the detection data;

The terrain parameter of the omnidirectional scanning area is determined according to the fitting plane, and the terrain parameter includes the slope of the omnidirectional scanning area.

In a third aspect, this application provides a radar, the radar including a memory and a processor;

The memory is used to store a computer program;

The processor is configured to execute the computer program and, when executing the computer program, implement the following steps:

Obtain the detection data of the omnidirectional scanning area obtained by the radar during the ground rotation scanning process;

Fitting a fitting plane of the omnidirectional scanning area according to the detection data;

The terrain parameter of the omnidirectional scanning area is determined according to the fitting plane, and the terrain parameter includes the slope of the omnidirectional scanning area.

In a fourth aspect, the present application provides a method for controlling drone operations. The drone is equipped with radar and operates on slopes. The method includes:

Acquiring the omnidirectional scanning area detection data obtained by the radar during the ground rotation scanning process;

Fitting a fitting plane of the omnidirectional scanning area according to the detection data;

Determining the terrain parameter of the omnidirectional scanning area according to the fitting plane, the terrain parameter including the slope of the omnidirectional scanning area;

Adjust the flight action of the drone according to the terrain parameter.

In a fifth aspect, this application provides an unmanned aerial vehicle, the unmanned aerial vehicle including a memory and a processor;

The memory is used to store a computer program;

The processor is configured to execute the computer program and, when executing the computer program, implement the following steps:

Acquiring the omnidirectional scanning area detection data obtained by the radar during the ground rotation scanning process;

Fitting a fitting plane of the omnidirectional scanning area according to the detection data;

Determining the terrain parameter of the omnidirectional scanning area according to the fitting plane, the terrain parameter including the slope of the omnidirectional scanning area;

Adjust the flight action of the drone according to the terrain parameter.

In the sixth aspect, this specification provides a computer-readable storage medium, the computer-readable storage medium stores a computer program, and the computer program can be used by a processor to implement the above-mentioned method.

The embodiments of this specification provide a terrain prediction method, device, radar, unmanned aerial vehicle and operation control method, computer-readable storage medium for slope land According to the detection data of the scanning area, the fitting plane of the omnidirectional scanning area is fitted according to the detection data; to realize the determination of the terrain parameters of the omnidirectional scanning area according to the fitting plane; and according to the terrain parameters Adjust the flight action of the drone. Since the omnidirectional scanning area includes the different directions of the UAV, the terrain parameters obtained are more global and accurate, so that the flight movements of the UAV can be controlled more safely.

It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the disclosure of this specification.

Description of the drawings

In order to explain the technical solutions of the embodiments of this specification more clearly, the following will briefly introduce the drawings used in the description of the embodiments. Obviously, the drawings in the following description are some embodiments of this specification. Ordinary technicians can obtain other drawings based on these drawings without creative work.

FIG. 1 is a schematic flowchart of a method for controlling drone operations according to an embodiment of this specification;

Figure 2 is a schematic diagram of the structure of an unmanned aerial vehicle according to an embodiment of this specification;

Figure 3 is a schematic structural diagram of a radar provided by an embodiment of this specification, in which the housing is not shown;

Figure 4 is a cross-sectional view of the radar provided by an embodiment of the specification, in which the housing is not shown;

FIG. 5 is a schematic diagram of the intersection of a rotation axis and a preset plane provided by an embodiment of the present application; FIG.

Fig. 6 is a schematic diagram of the structure of a radar according to an embodiment of the present specification;

FIG. 7 is a schematic structural diagram of a terrain prediction device according to an embodiment of this specification;

FIG. 8 is a schematic diagram of the radar scanning the omnidirectional scanning area during the rotation process in the embodiment of this specification;

Figure 9 is a schematic diagram of radar scanning a ground point to obtain detection data;

Figure 10 is a schematic diagram of the distribution of several ground points in the geodetic coordinate system in the omnidirectional scanning area;

FIG. 11 is a schematic diagram of the fitting plane of the omnidirectional scanning area;

Fig. 12 is a schematic block diagram of an unmanned aerial vehicle according to an embodiment of the present specification.

Detailed ways

The technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by a person of ordinary skill in the art without creative work shall fall within the protection scope of this specification.

The flowchart shown in the drawings is only an example, and does not necessarily include all contents and operations/steps, nor does it have to be executed in the described order. For example, some operations/steps can also be decomposed, combined or partially combined, so the actual execution order may be changed according to actual conditions.

Hereinafter, some embodiments of this specification will be described in detail with reference to the accompanying drawings. In the case of no conflict, the following embodiments and features in the embodiments can be combined with each other.

Please refer to FIG. 1. FIG. 1 is a schematic flowchart of a method for controlling drone operations according to an embodiment of this specification.

The drone operation control method can be applied to the drone to control the drone's flight actions and other processes according to the terrain. Among them, the drone can be a rotary-wing drone, such as a four-rotor drone, a six-rotor drone, an eight-rotor drone, or a fixed-wing drone.

Fig. 2 is a schematic diagram of the structure of an unmanned aerial vehicle in an embodiment. The embodiment of this specification takes the rotary wing unmanned aerial vehicle as an example for description.

Referring to FIG. 2, an embodiment of the present application provides an unmanned aerial vehicle 1000. The unmanned aerial vehicle 1000 may include a body 100, a spraying mechanism 220, a power system 300, and a flight control system. The UAV 1000 can communicate with the control terminal wirelessly. The control terminal can display the flight information of the UAV 1000, etc. The control terminal can communicate with the UAV 1000 wirelessly for remote control of the UAV 1000. .

Wherein, the airframe 100 may include a fuselage 110 and a landing gear 120. The fuselage 110 may include a center frame 111 and one or more arms 112 connected to the center frame 111, and the one or more arms 112 extend radially from the center frame 111. The landing frame 120 is connected to the fuselage 110 and is used for supporting the UAV 1000 when it is landed.

In some embodiments, the spraying mechanism 220 is provided on the body 110, and the spraying mechanism 220 is connected to the containing box 210 for spraying the objects to be sprayed in the containing box 210. The object to be sprayed can be liquid medicine, water or fertilizer. Specifically, referring to FIG. 2 again, the spraying mechanism 220 includes a water pump and a spray head 221. The accommodating box 210 is used to store liquid medicine or water. The storage box 210 and the water pump are mounted on the body 110. The spray head 221 is mounted on the end of the arm 112. The liquid in the containing box 210 is pumped into the spray head 221 by a water pump, and sprayed out by the spray head 221. The power system 300 can drive the body 100 to move, rotate, turn, etc., so as to drive the spray head 221 to move to different positions or different angles to perform spraying operations in a preset area.

The power system 300 may include one or more electronic governors (referred to as ESCs for short), one or more propellers 310, and one or more power motors 320 corresponding to the one or more propellers 310, wherein the power motors 320 are connected Between the electronic governor and the propeller 310, the power motor 320 and the propeller 310 are arranged on the arm 112 of the UAV 1000; the electronic governor is used to receive the driving signal generated by the flight control system and provide a driving current according to the driving signal The power motor 320 is provided to control the speed of the power motor 320. The power motor 320 is used to drive the propeller 310 to rotate, so as to provide power for the flight of the drone 1000, and the power enables the drone 1000 to realize one or more degrees of freedom of movement. In some embodiments, the drone 1000 may rotate around one or more rotation axes. For example, the aforementioned rotation axis may include a roll axis (roll axis), a yaw axis (yaw axis), and a pitch axis (pitch axis). In some embodiments, the roll axis is the Y axis in FIG. 2, the pitch axis is the X axis in FIG. 2, and the yaw axis is the Z axis in FIG. 2. It should be understood that the power motor 320 may be a DC motor or an AC motor. In addition, the power motor 320 may be a brushless motor or a brushed motor.

The flight control system may include a flight controller and a sensing system. The sensing system is used to measure the attitude information of the unmanned aerial vehicle 1000, that is, the position information and state information of the unmanned aerial vehicle 1000 in space, such as three-dimensional position, three-dimensional angle, three-dimensional velocity, three-dimensional acceleration, and three-dimensional angular velocity. The sensing system may include, for example, at least one of sensors such as a gyroscope, an ultrasonic sensor, an electronic compass, an inertial measurement unit (IMU), a vision sensor, a global navigation satellite system, and a barometer. For example, the global navigation satellite system may be the Global Positioning System (GPS). The flight controller is used to control the flight of the drone 1000, for example, it can control the flight of the drone 1000 according to the attitude information measured by the sensor system. It should be understood that the flight controller can control the drone 1000 according to pre-programmed program instructions, and can also control the drone 1000 by responding to one or more control instructions from the control terminal.

As shown in Fig. 2, the landing frame 120 of the UAV 1000 is equipped with a radar 400, which can detect objects, such as obstacles. Specifically, the radar 400 can measure the distance from the object to the launch point of the radar 400, the rate of change of distance, the azimuth, the height, etc., so as to realize functions such as ground scanning. In some embodiments, the radar 400 is a millimeter wave radar 400. Of course, in other embodiments, the radar 400 may also be an over-the-horizon radar 400, a microwave radar 400, a lidar 400, or the like.

Please refer to FIG. 3 and FIG. 4, where the radar 400 includes a base 410, an antenna mechanism 420 and a driving mechanism 430. The antenna mechanism 420 can rotate relative to the fuselage 110 around a preset rotation axis R, and is used to detect obstacles on the side of the UAV 1000.

In some embodiments, the base 410 is mounted on the landing gear 120. The antenna mechanism 420 includes a transmitter (not labeled) and a receiver (not labeled). The transmitter is used to generate and transmit radar signals. The radar signals propagate forward in the transmitted direction and are reflected when they encounter obstacles. The receiver is used to receive the echo signal that is reflected back.

The antenna mechanism 420 can rotate around the rotation axis R under the driving of the driving mechanism 430, so that the antenna mechanism 420 can selectively transmit signals in multiple directions and receive echo signals reflected from multiple directions. Therefore, the distance between the drone 1000 and obstacles in multiple directions can be selectively detected by one antenna mechanism 420, and the structure of the drone 1000 is simple.

In some embodiments, the rotation axis R intersects the preset plane ω, that is, the rotation axis R and the preset plane ω are arranged non-parallel. The preset plane ω is the plane where the pitch axis and roll axis of the drone 1000 are located. As a result, the radar 400 can not only detect the front and rear fields of view of the UAV 1000 to scan the front and rear, but also can detect other side fields of view of the UAV 1000 in addition to the front field of view and the rear field of view, expanding the unmanned field. The detection angle and detection coverage of the machine 1000 ensure the omnidirectionality of scanning.

In some embodiments, the driving mechanism 430 is provided on the base 410. The rotating part of the driving mechanism 430 is connected to the antenna mechanism 420 to drive the antenna mechanism 420 to rotate around the rotation axis R. Specifically, the driving mechanism 430 includes a motor, and the motor includes a stator 431 and a rotor 432. The rotor 432 is a rotating part of the driving mechanism 430. The rotor 432 can rotate relative to the stator 431 to drive the antenna mechanism 420 to rotate. More specifically, the stator 431 is mounted on the base 410, and the antenna mechanism 420 is mounted on the rotor 432 of the motor. The rotor 432 rotates relative to the base 410, so that the antenna mechanism 420 rotates about the rotation axis R relative to the base 410.

Specifically, the antenna mechanism 420 of the radar 400 is driven by the rotor 432 to rotate in the forward or reverse direction around the rotation axis R based on the nose direction of the drone 1000, and scans a fan-shaped area within an angular range each time. The antenna mechanism 420 rotates one circle, that is, 360°, and can scan a complete circular area with the center of the radar 400 as the center, so as to obtain the detection data of the circular omnidirectional scanning area.

In some embodiments, the rotor 432 of the motor can rotate forward or backward for at least one turn, thereby driving the antenna mechanism 420 to rotate forward or backward at least 360° in all directions. Specifically, the rotation angle range of the antenna mechanism 420 around the rotation axis R is greater than or equal to 360°, such as 450°, 540°, 720°, 1020°, etc., to achieve continuous rotation, thereby increasing the data collection points of the antenna mechanism 420 and improving the radar 400 measurement accuracy.

In some embodiments, please refer to FIG. 5, the angle α between the rotation axis R and the preset plane ω is 60°-90°. Specifically, the included angle α between the rotation axis R and the preset plane ω may be 60°, 65°, 70°, 80°, 85°, 90°, and any other suitable angles between 60° and 90°. The angle α between the rotation axis R and the preset plane ω is in the range of 60°-90°, so that the scanning field of view can not only include the front field of vision and the rear field of vision, but also include other side field of vision except the front field of vision and the rear field of vision. , So as to expand the detection angle and detection coverage of UAV 1000 as much as possible, and realize omnidirectional scanning.

In some embodiments, the rotation axis R roughly coincides with the center line of the fuselage 110 to avoid the problem of unbalanced center of gravity of the UAV 1000 caused by the installation of the radar 400, thereby ensuring the flight reliability of the UAV 1000. Wherein, substantially coincident means that the angle between the rotation axis R and the center line of the fuselage 110 is 0°-10°, that is, any angle between 0°, 10°, and 0°-10°.

In some embodiments, the rotation axis R and the yaw axis of the drone 1000 are at an acute angle. Wherein, the acute angle can be any suitable angle, for example, 0°-30°, that is, any of 0°, 5°, 10°, 15°, 20°, 25°, 30°, and 0°-30° Other suitable angles.

In some embodiments, the rotation axis R is substantially perpendicular to the preset plane ω, or the rotation axis R is substantially coincident with the yaw axis of the UAV 1000. At this time, the omnidirectional scanning area of the radar 400 is a circle centered on the center of the radar 400. The perfect circle is a 360° area surrounding the side of the UAV 1000, which can reflect the ground detection information of the UAV 1000 in different directions.

When the rotation axis R of the rotor 432 of the driving mechanism 430 is perpendicular to the preset plane ω, that is, when the rotation axis R of the rotor 432 is perpendicular to the plane where the pitch and roll axes of the drone 1000 are located, by adjusting the rotation angle of the antenna mechanism 420, The antenna mechanism 420 can transmit microwave signals to the left, right, front, and rear of the UAV 1000 and receive echo signals reflected by obstacles on the left, right, front, and rear. At this time, the radar 400 is available It is used to realize the functions of left scan, right scan, front scan, rear scan, left terrain prediction, right terrain prediction, front terrain prediction, and rear terrain prediction. Of course, the intersection of the rotation axis R of the rotor 432 with the plane where the pitch axis and roll axis of the UAV 1000 are located can also be other specific situations, which are not limited here.

It is understandable that there is a preset angle between the rotation axis R and the preset plane ω, or when the rotation axis R and the yaw axis of the UAV 1000 are at an acute angle, the omnidirectional scanning area is not a perfect circle, but it is also a circle. The 360° area of the UAV 1000 can reflect the ground detection information of the UAV 1000 in different directions.

It should be noted that the above-mentioned rotation axis R may be a real axis or an imaginary axis. When the rotation axis R is a real axis, the antenna mechanism 420 can rotate relative to the rotation axis R; or, the antenna mechanism 420 rotates along with the rotation axis R.

In some embodiments, the antenna mechanism 420 is provided on the side of the base 410 away from the fuselage 110, so that the antenna mechanism 420 of the radar 400 is as far away from the sensor provided on the fuselage 110 as possible, and the radar generated by the antenna mechanism 420 is reduced. Signals (such as electromagnetic waves) interfere with sensors on the body 110.

Please refer to FIG. 3 and FIG. 4. In some embodiments, the radar 400 further includes a sensing mechanism 440. The sensing mechanism 440 is arranged at an end of the antenna mechanism 420 far away from the base 410 and is used to detect the height of the drone 1000 relative to the ground. When the driving mechanism 430 drives the antenna mechanism 420 to rotate, the sensing mechanism 440 also rotates together with the antenna mechanism 420. The sensing mechanism 440 includes at least one of a vision sensor, an ultrasonic ranging sensor, a depth camera, a radar antenna structure, and the like.

It can be understood that the shape of the antenna mechanism 420 and the sensing mechanism 440 can be designed in any suitable shape according to actual requirements, for example, a plate shape. Exemplarily, when the antenna mechanism 420 and the sensing mechanism 440 are both substantially plate-shaped, the antenna mechanism 420 and the sensing mechanism 440 are substantially perpendicular. Specifically, the antenna mechanism 420 is substantially perpendicular to the plane where the pitch axis and the roll axis of the drone 1000 are located. The sensing mechanism 440 is substantially parallel to the plane where the pitch axis and roll axis of the drone 1000 are located.

Please refer to FIG. 3 and FIG. 4. In some embodiments, the radar 400 further includes a circuit board 450. The circuit board 450 is disposed on the base 410 opposite to the antenna mechanism 420 for processing the signal of the antenna mechanism 420. Specifically, the circuit board 450 can process the signal of the antenna mechanism 420, for example, amplify the echo signal; filter the interference signal; convert the echo signal into a radar data signal for the control of the back-end equipment, terminal observation and/or recording Wait.

In some embodiments, the circuit board 450 has a plate shape, but of course it can also be designed in any other suitable shape. Since the center of gravity of the antenna mechanism 420 deviates from the rotation axis R of the antenna mechanism 420, the center of gravity of the radar 400 will deviate from the rotation axis R of the antenna mechanism 420, thereby causing the center of gravity of the drone 1000 to be unbalanced, making the drone 1000 unreliable in flight. To this end, the circuit board 450 and the antenna mechanism 420 are arranged at opposite ends of the sensing mechanism 440. The circuit board 450 and the antenna mechanism 420 are arranged symmetrically about the rotation axis R, so as to balance the center of the antenna mechanism 420 so that the center of the radar 400 is approximately located The rotation axis R of the antenna mechanism 420 is mounted. Specifically, the antenna mechanism 420, the sensing mechanism 440, and the circuit board 450 form a “Π” structure, and the opening of the “Π” structure faces the body 110.

Please refer to FIG. 2 again. In some embodiments, the radar 400 further includes a housing 460. The housing 460 cooperates with the base 410 to form an accommodating space. The antenna mechanism 420, the driving mechanism 430, the sensing mechanism 440 and the circuit board 450 are accommodated in the housing. In the space, the antenna mechanism 420, the driving mechanism 430, the sensing mechanism 440, and the circuit board 450 are protected from the external environment, so as to avoid interference from the external environment or damage to these components. It is understandable that the signals transmitted or received by the antenna mechanism 420 and the sensing mechanism 440 can be passed through the housing 460, that is, the housing 460 will not affect the normal transmission or reception of signals by the antenna mechanism 420 and the sensing mechanism 440.

Because the antenna mechanism of the radar can rotate around the axis of rotation, the axis of rotation intersects the plane where the pitch axis and the roll axis are located, not only can detect the front and rear vision of the UAV, but also the side vision of the UAV in addition to the front and rear vision The detection is carried out on the side field of view other than the field of view, which expands the detection angle and detection coverage of the UAV, and realizes the omnidirectional scanning of the ground.

It is understandable that the above-mentioned naming of the components of the drone is only for identification purposes, and should not be understood as a limitation to the embodiments of this specification.

As shown in Fig. 1, the operation control method of the drone in this embodiment includes steps S110 to S140.

S110. Obtain the detection data of the omnidirectional scanning area obtained by the radar during the ground rotation scanning process.

Exemplarily, the radar is mounted under the UAV, and the detection data of the circular omnidirectional scanning area with the center of the circle directly under the radar can be obtained.

S120: Fit a fitting plane of the omnidirectional scanning area according to the detection data.

Exemplarily, the detection data includes the azimuth information of several ground points in the omnidirectional scanning area, and the fitting plane of the ground in the omnidirectional scanning area is obtained by fitting, and most of the ground points in the omnidirectional scanning area are located on the fitting plane or The distance to the fitting plane is small.

S130. Determine the terrain parameter of the omnidirectional scanning area according to the fitting plane.

Exemplarily, when the drone is operating on a slope, it is more necessary to pay attention to the slope of the omnidirectional scanning area, so the determined terrain parameters may include the slope.

Exemplarily, information such as the slope of the omnidirectional scanning area can be determined according to the inclination direction of the fitting plane.

Exemplarily, when the drone is operating on undulating ground, it is more necessary to pay attention to the flatness of the omnidirectional scanning area, so the determined terrain parameters may include the flatness.

Exemplarily, it can be determined whether the omnidirectional scanning area is flat according to the number of ground points and the size of the distance with a larger distance from the fitting plane.

It is understandable that, in some embodiments, the aforementioned steps S110 to S130 may be implemented by a drone.

It is understandable that, in some implementations, the foregoing steps S110 to S130 can be implemented by a radar, that is, the steps of the slope terrain prediction method in the embodiment of this specification.

Exemplarily, as shown in FIG. 6, the radar 400 includes a processor 401 and a memory 402.

For example, referring to FIGS. 3 and 4, the processor 401 and the memory 402 may be provided on the circuit board 450 of the radar 400.

Exemplarily, the processor 401 and the memory 402 are connected by a bus 403, and the bus 403 is, for example, an I2C (Inter-integrated Circuit) bus.

Specifically, the processor 401 may be a micro-controller unit (MCU), a central processing unit (CPU), a digital signal processor (Digital Signal Processor, DSP), or the like.

Specifically, the memory 402 may be a Flash chip, a read-only memory (ROM, Read-Only Memory) disk, an optical disk, a U disk, or a mobile hard disk.

The memory 402 is used to store computer programs;

The processor 401 is configured to execute the computer program and, when executing the computer program, implement the aforementioned steps S110 to S130, that is, the steps of the slope terrain prediction method in the embodiment of this specification.

It is understandable that, in some embodiments, the aforementioned step S110 to step S130 can be implemented by the terrain prediction device, that is, the steps of the slope terrain prediction method in the embodiment of this specification.

The terrain prediction device may be, for example, a server or a terminal. Among them, the terminal can be a mobile phone, a tablet computer, a notebook computer, a desktop computer, a personal digital assistant, and other electronic devices; the server can be an independent server or a server cluster.

Exemplarily, as shown in FIG. 7, the terrain prediction apparatus 600 includes a processor 601 and a memory 602.

Exemplarily, the processor 601 and the memory 602 are connected by a bus 603, and the bus 603 is, for example, an I2C (Inter-integrated Circuit) bus.

Specifically, the processor 601 may be a micro-controller unit (MCU), a central processing unit (CPU), a digital signal processor (Digital Signal Processor, DSP), or the like.

Specifically, the memory 602 may be a Flash chip, a read-only memory (ROM, Read-Only Memory) disk, an optical disk, a U disk, or a mobile hard disk.

The memory 602 is used to store computer programs;

The processor 601 is configured to execute the computer program and, when executing the computer program, implement the aforementioned steps S110 to S130, that is, the steps of the slope terrain prediction method in the embodiment of this specification.

The slope terrain prediction method, device, and radar provided in the above-mentioned embodiments of this specification acquire the detection data of the omnidirectional scan area obtained by the radar during the ground rotation scanning process, and fit the omnidirectional scan area according to the detection data The fitting plane; in order to realize the determination of the terrain parameters of the omnidirectional scanning area according to the fitting plane, it can be based on the detection of the UAV's various positions, such as the front, rear, left, and right ground, for a more comprehensive Predict the terrain of the area where the drone is located.

In some embodiments, the drone may also perform step S140.

S140. Adjust the flying action of the drone according to the terrain parameter.

For example, the drone can obtain terrain parameters from a radar, obtain terrain parameters from a terrain prediction device, or obtain the terrain parameters by the drone through the aforementioned steps S110 to S130.

Exemplarily, the UAV can adjust the flight actions of the UAV according to the information such as the flatness and slope of the ground, so as to ensure the safe flight of the UAV and the reliable execution of the operation tasks.

In some embodiments, obtaining the omnidirectional scanning area detection data obtained by the radar during the ground rotation scanning process in step S110 includes: obtaining the radar's return data, and performing spectrum extraction, processing and analysis on the return data , You can calculate the relative spatial position between the radar and the scanning target, such as obstacles.

Exemplarily, the radar is a continuous wave radar. The detection data can be obtained according to the continuous wave radar ranging and angle measurement algorithm. For example, by processing the radar return data, spectrum extraction and further spectrum refinement are completed, and finally the refined frequency point position is converted into the azimuth information of the ground point.

Exemplarily, as shown in FIG. 8, the radar mounted under the drone rotates based on the direction of the drone's nose, and scans a fan-shaped area within an angular range each time. The radar rotates one circle, that is, 360°, it can scan a complete circular area, and obtain the detection data of the circular omnidirectional scanning area with the center of the circle directly below the radar.

Exemplarily, the rotation direction of the radar can be the same as the vertical direction of the drone. At this time, the omnidirectional scanning area is a perfect circle with the center of the radar as the center, which is a 360° area surrounding the drone. It reflects the ground detection information of the UAV in different directions.

Exemplarily, there may be a preset angle between the rotation direction of the radar and the vertical direction of the drone. At this time, the omnidirectional scanning area is not a perfect circle, but it is also a 360° area surrounding the drone. It can reflect the ground detection information of the UAV in front, back, left, and right directions.

In this embodiment, a certain area under the drone is scanned based on a rotating radar to obtain the spatial orientation information of the ground point in the omnidirectional scanning area.

In some embodiments, the detection data of several ground points in the omni-directional scanning area can also be acquired through a vision sensor, a time of flight (TOF) sensor, or a sensor module with ranging and angle measurement such as a lidar or an ultrasonic module. For example, a two-dimensional image of the omnidirectional scanning area is acquired by a vision sensor, and then a three-dimensional point cloud is extracted from the two-dimensional image. However, the visual sensor has higher requirements for the lighting environment, and is easily affected by the light intensity, the color of the background target, the dust in the environment, and the water mist.

In some implementation manners, the acquiring the omnidirectional scanning area detection data obtained by the radar during the ground rotation scanning process includes: acquiring the detection distances and azimuth angles of several ground points in the omnidirectional scanning area relative to the radar .

Figure 9 shows the radar detection target object, such as a schematic diagram of the beam horizontal plane of a certain ground point. Exemplarily, the output of the radar is the detection distance r and the azimuth angle θ of the ground point relative to the radar.

Specifically, the detection distance r of the ground point relative to the radar represents the radial distance of the ground point relative to the center of the radar.

Exemplarily, the method further includes: acquiring the radar rotation angle corresponding to each of the ground points when the radar detects each of the ground points

For example, the radar rotation angle

Indicates the relative initial position of the radar's radio frequency board at the current frame, such as the rotation angle of the nose.

It can be understood that in step S110, the coordinate system of several ground points in the omnidirectional scanning area with the radar as the origin can be obtained, which can be called the coordinates on the radar coordinate system. The radar coordinate system can be, for example, a spherical coordinate system, a cylindrical coordinate system, etc. , Of course, it can also be a rectangular coordinate system.

The coordinates of the ground point on the rectangular coordinate system with the radar as the origin may be expressed in terms of the distance of the ground point relative to the radar in multiple directions.

Exemplarily, the spherical coordinate system coordinates, the cylindrical coordinate system coordinates, etc., of the ground point obtained from the radar may be converted to the coordinates on the rectangular coordinate system. It is convenient for plane fitting calculation, the calculation amount is smaller, and the terrain can be judged faster.

Exemplarily, the acquiring the omnidirectional scanning area detection data obtained by the radar during the ground rotation scanning process further includes: determining each of the ground points according to the radar rotation angle, detection distance, and azimuth angle corresponding to the ground point The distance in multiple directions relative to the radar.

It is understandable that the acquiring the detection data of the omnidirectional scanning area obtained by the radar during the ground rotation scanning process may include: acquiring the distances of several ground points in the omnidirectional scanning area relative to the radar in multiple directions .

Specifically, the distances of the ground point relative to the radar in multiple directions include: position coordinates of the ground point in the radar coordinate system of the radar.

Wherein, the radar coordinate system takes the rotation center of the radar as the origin, the first axis direction is directly below the radar, the second axis direction is the front direction of the radar, and is perpendicular to the first axis direction. The direction of the uniaxial direction and the second axis direction is the third axis direction.

Exemplarily, the position coordinates of the ground point in the radar coordinate system of the radar are represented by {x, y, z} _A , which is specifically represented by the Cartesian coordinate system of the radar observation system of the ground point, where the subscript A represents the coordinate system The origin of is determined according to the center of the radar, and x, y, and z respectively represent three mutually perpendicular coordinate axes of the Cartesian coordinate system.

Exemplarily, the spherical coordinate system coordinates of the ground point are

Representation, the conversion of the ground point position representation from the spherical coordinate system to the Cartesian coordinate system can be accomplished by the following model:

In some embodiments, in step S110, coordinate data of several ground points in the omnidirectional scanning area in the geodetic coordinate system can be acquired.

Specifically, the geodetic coordinate system takes a certain point on the ground as its origin. For example, the origin of the geodetic coordinate system is located directly below the radar. In order to obtain the fitting surface of the omnidirectional scanning area under the radar in subsequent fitting.

The geodetic coordinate system takes the true north or true south direction of the geodetic origin as the fourth axis direction, takes the true east or west of the geodetic origin as the fifth axis direction, and is perpendicular to the fourth axis and the direction. The direction of the fifth axis direction is the sixth axis direction.

Exemplarily, the coordinate data of the ground point in the geodetic coordinate system includes the distance of the ground point in the fourth axis direction, the fifth axis direction, and the sixth axis direction relative to the geodetic origin.

Exemplarily, the adopted geodetic coordinate system is ENU (East-North-UP coordinate system). _{X G} can be used to indicate the distance of the ground point relative to the origin of the coordinate from the true north, y _G means the distance of the ground point from the origin of the coordinate to the true east, and z _G the distance of the ground point from the origin of the coordinate in the vertical direction.

Specifically, the radar can convert the coordinates of the ground point on the coordinate system with the radar as the origin to the coordinates on the geodetic coordinate system according to the conversion model between the coordinate system with the radar as the origin and the geodetic coordinate system. The conversion model can be determined based on the attitude of the radar, for example.

In some embodiments, the UAV can obtain the coordinates of the ground point in the coordinate system with the radar as the origin from the radar, and obtain the attitude of the radar, and then the UAV obtains the ground point in the geodetic coordinate system according to the attitude of the radar. coordinate of.

Exemplarily, step S120 fitting the fitting plane of the omnidirectional scanning area according to the detection data includes: determining the coordinates of several ground points in the omnidirectional scanning area in the geodetic coordinate system according to the detection data Data; fitting the fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points.

By converting the ground point from the radar observation system to the geodetic coordinate system, the influence of radar or radar carrier, such as the attitude of the UAV, on the ground observation can be eliminated, and the plane model obtained by the ground fitting is more accurate.

Exemplarily, the conversion of the coordinates of the ground point on the coordinate system with the radar as the origin to the coordinates on the geodetic coordinate system may be converted by the calculation model of the following formula:

among them,

Represents the homogeneous transformation matrix between the coordinate system with the radar as the origin and the geodetic coordinate system, and G represents the geodetic coordinate system.

In some embodiments, the method further includes: acquiring the attitude information of the radar through an inertial measurement unit (IMU) carried by the drone and/or an inertial measurement unit carried by the radar.

Exemplarily, the determining the coordinate data of several ground points in the omnidirectional scanning area in the geodetic coordinate system according to the detection data includes: determining the position of the several ground points according to the attitude information of the radar and the detection data Coordinate data in the geodetic coordinate system.

_{Exemplarily, the attitude quaternion {q 0} , q ₀ , q ₀ , q ₀ } of the radar is obtained from the inertial measurement unit in real time, and the coordinate system with the radar as the origin and the geodetic coordinate system can be determined according to the attitude quaternion of the radar The homogeneous transformation matrix between.

Exemplary, there are:

among them,

Is the rotation matrix determined according to the attitude quaternion of the radar,

Is a constant vector.

Through the above preprocessing, the coordinate data of several ground points in the omnidirectional scanning area in the geodetic coordinate system can be obtained. As shown in Figure 10, the ground points can be expressed as {x _G , y _G , z _G }.

In some embodiments, fitting the fitting plane of the omnidirectional scanning area according to the detection data in step S120 includes: fitting the omnidirectional scanning area according to the coordinate data of the several ground points The fitted plane of the scan area.

In some embodiments, fitting the fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points in step S120 includes: filtering the coordinate data of the several ground points, and according to the filtering The coordinate data of the subsequent ground point fits the fitting plane of the omnidirectional scanning area.

Due to the interference of the internal and external environments of the radar, there will be outliers in the distance measured by the radar. For example: For the same ranging point, the distance between the ranging point and the radar is actually relatively large, but the radar is interfered, resulting in smaller data obtained from the ranging, which in turn will lead to the fitted plane and the predicted terrain There is a large error in the parameters. Especially in complex application scenarios such as farmland and tea mountains, the presence of outliers will lead to inaccurate terrain predictions.

The outlier points and object points existing in the original radar observations will make the assumption that the ground is a flat plane invalid. If the plane fitting is directly performed on the coordinate data of the ground point, the fitting result will have a large deviation from the actual situation.

In order to make the fitting plane of the omnidirectional scanning area closer to the actual plane, outliers and ground object points can be eliminated before plane fitting. Among them, the feature point represents the ground attachment, such as the ground point of a building, which is not the real ground.

Therefore, by eliminating outliers and non-ground points, and then fitting the fitting plane of the omnidirectional scanning area according to the coordinate data of the unremoved ground points, accurate ground estimation can be achieved.

In some embodiments, first perform a cluster analysis on the several ground points according to the coordinate data of the several ground points to determine the ground points that meet the clustering conditions; and then fit the ground points according to the coordinate data of the ground points that meet the clustering conditions The fitting plane of the omnidirectional scanning area is obtained.

Exemplarily, the DBSCAN clustering algorithm can be used to eliminate outliers and ground feature points in the original observations, and extract effective ground points.

Exemplarily, the performing cluster analysis on the plurality of ground points according to the coordinate data of the plurality of ground points to determine the ground points that meet the clustering condition includes:

Repeat the following steps until the several ground points are determined as points to be clustered: determine one of the ground points as points to be clustered, and determine the number of ground points within the search range of the points to be clustered If the number is not greater than the preset clustering threshold, the points to be clustered are eliminated; if the loop is stopped, it is determined that the unremoved points to be clustered are ground points that meet the clustering condition.

For example, use the KDTREE data structure to establish a tree structure for all ground points based on the coordinate data of the ground points; then determine a ground point in turn as the point to be clustered, and find all the points in the tree that are less than the preset search Radius of ground points, if the number of ground points within the search radius of the points to be clustered is greater than the preset minimum point cluster threshold, the current points to be clustered are considered as valid points and reserved, otherwise, they are considered as outliers or miscellaneous points. Point elimination; until each ground point in the omnidirectional scanning area is traversed.

It is understandable that the coordinate data of several ground points in the omni-directional scanning area can also be screened in other ways to eliminate outliers and noise points in the original observation, for example, through filtering methods such as feature segmentation.

In some embodiments, the fitting plane of the omnidirectional scanning area is fitted according to the coordinate data of the ground point, and the plane fitting may be performed by a least square method. For example, the fitting plane of the omnidirectional scanning area is fitted by the least square method according to the coordinate data of the filtered ground points.

Exemplarily, the vertical height z is regarded as the variable with the highest degree of independence, and the plane equation is established:

Z=aX+bY+c

Then the coordinate data of the ground point can be fitted by the least square method. The least square method, also known as the least square method, is a mathematical optimization method. It finds the best function match of the data by minimizing the sum of squares of the error. The least square method can be used to easily obtain unknown data, and minimize the sum of squares of errors between the obtained data and the actual data. The "least squares method" is a standard method to obtain approximate solutions for overdetermined equations, that is, equations with more equations than unknowns, by regression analysis. In this entire solution, the least squares method is calculated as the result of each equation, minimizing the sum of the residual sum of squares.

For example, using Cramer's rule to determine the parameters of the plane equation:

Among them, the coordinates of the target center point can be based on

_{The coordinates {x G} , y _G , z _G } of the ground points participating in the fitting are determined by the average value,

Is the coordinates of the ground point {x _G , y _G , z _G } minus the target center point

The normalized result obtained.

Therefore, the fitting plane Z=aX+bY+c of the omnidirectional scanning area can be fitted according to the coordinate data of the ground point.

In other embodiments, the effective points in the original observation can be screened out by the least squares method based on RANSAC, that is, the coordinate data of the several ground points are screened to obtain the coordinate data of the ground points after screening.

Random sampling consensus algorithm (RANdom SAmple Consensus, RANSAC) uses an iterative method to estimate the parameters of the mathematical model from a set of observed data containing outliers. RANSAC is a non-deterministic algorithm. In a sense, it will produce a reasonable result with a certain probability, and more iterations will increase this probability.

Exemplarily, the fitting the fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points includes: determining at least three ground points from the several ground points, and according to the at least three ground points. Ground points determine a target plane; determine the plane distance from each ground point to the target plane according to the coordinate data of the several ground points; if the plane distance is not greater than the distance threshold, the number of ground points is not less than the preset The number threshold is fitted to obtain the fitted plane of the omnidirectional scanning area according to the ground points whose plane distance is not greater than the distance threshold.

Exemplarily, the fitting of the fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points includes the following steps:

The first step: randomly select at least three non-collinear ground points (x ₁ ,y ₁ ,z ₁ ), (x ₂ ,y ₂ ,z ₂ ), (x ₃ ,y ₃ ,z ₃ ); and establish a plane passing through the at least three points, for example: aX+bY+Cz+D=0, where:

a=(y ₂ -y ₁ )×(z ₃ -z ₁ )-(y ₃ -y ₁ )×(z ₂ -z ₁ );

b=(z ₂ -z ₁ )×(x ₃ -x ₁ )-(z ₃ -z ₁ )×(x ₂ -x ₁ );

c=(x ₂ -x ₁ )×(y ₃ -y ₁ )-(x ₃ -x ₁ )×(y ₂ -y ₁ );

D=a×x ₁ -b×y ₁ -c×z ₁ .

The second step: Calculate _{the distance between all ground points {x i} , y _i , z _i } and the plane established in the first step:

The third step: if the distance from a certain ground point to the plane is less than a preset threshold, the ground point is considered to be an intra-office point, and the number n of intra-office points corresponding to the plane is counted.

Fourth step: If the number n of intra-office points is greater than the number N of intra-office points required to establish a credible plane (preset value), use all the intra-office points corresponding to the plane to perform plane fitting to obtain the fitted plane. For example, fitting by the least squares method.

Exemplarily, the fitting of the fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points further includes: according to the height value of the interior point, such as z ₃ and the intercept D of the fitting plane It is judged whether the fitting plane obtained by fitting the interior points is accurate.

Exemplarily, if the difference between the intercept D of the fitting plane and the height values of all interior points is less than a preset threshold, the fitting plane obtained by fitting the interior points is accurate, and the fitting result of this time can be retained.

Exemplarily, if the difference between the intercept D of the fitting plane and the height values of all interior points is not less than a preset threshold, then the fitting plane fitted this time is discarded. It is possible to return to the determination of at least three ground points from the plurality of ground points, and determine a target plane according to the at least three ground points; determine each ground point to the target according to the coordinate data of the plurality of ground points The plane distance of the plane; if the number of ground points whose plane distance is not greater than the distance threshold is not less than the preset number threshold, the omnidirectional scan area is obtained by fitting the ground points whose plane distance is not greater than the distance threshold The step of combining the plane is continued, and the at least three ground points are different from at least one of the three ground points determined last time.

Exemplarily, if the outliers are eliminated, that is, the number of ground points after screening is less than the number required to establish a credible plane, it is determined that the radar acquired this time has a certain number of omnidirectional scanning areas during the rotation process. The detection data of the ground point is invalid. Can re-acquire and filter, fit.

Exemplarily, the fitting plane of the omnidirectional scanning area is fitted in step S120 as shown in FIG. 11.

In some embodiments, the terrain parameters of the omnidirectional scanning area may be determined according to the fitting plane, and the terrain parameters of the omnidirectional scanning area include at least one of the following: the slope of the omnidirectional scanning area, the The flatness of the omnidirectional scanning area, and the height value of the ground directly below the radar.

In some embodiments, the drone operates on slopes, such as spraying pesticides on terraces, orchards on hillsides. At least part of the terrain has a certain angle relative to the horizontal, that is, the slope.

In this scenario, if only the detection data in front of the drone's nose is used to predict the changes in the terrain in front of the nose, the slope of the area where the drone is located cannot be fully reflected, such as the rear, left, and right sides of the drone. For the terrain on the side, the terrain following effect is poor, and the safety of drone operations is poor.

Exemplarily, the terrain parameter of the omnidirectional scanning area determined in step S130 according to the fitting plane includes the slope of the omnidirectional scanning area.

Exemplarily, the determining the terrain parameters of the omnidirectional scanning area according to the fitting plane includes: determining a slope direction; and determining the slope of the fitting plane in the slope direction.

For example, the slope of the fitting plane in the slope direction may be determined according to the normal vector of the fitting plane.

Exemplarily, it may be determined that the direction of the drone nose is the gradient direction; or it may be determined that the flying direction of the drone is the gradient direction. Of course, other directions can also be used as the gradient direction, for example, the left, right, and tail directions of the drone fuselage, or the left, right, etc. in the flying direction of the drone are determined as the gradient direction.

Exemplarily, the attitude of the radar, such as the rotation angle, is based on the direction of the nose.

Exemplarily, as shown in FIG. 11, if the direction of the nose is the gradient direction, or although the flying direction of the drone is the gradient direction but the flying direction of the drone is consistent with the nose, then the flight The slope of the direction can be directly determined by the normal vector of the fitted plane

get:

Exemplarily, if the determined gradient direction, for example, it is determined that the flying direction of the drone is the gradient direction, and the gradient direction is at a certain angle with the nose, the gradient of the fitting plane in the gradient direction can be obtained by the following formula :

Among them, {V _x , V _y } represents the gradient direction, for example, the projection of the flying direction of the drone on the fitting plane.

It is understandable that by acquiring the detection data of several ground points in the omnidirectional scanning area, and determining the fitting plane of the omnidirectional scanning area according to the detection data, the terrain slope in any direction of the omnidirectional scanning area can be determined according to the fitting plane. , Such as determining the slope of the ground in the flying direction of the drone.

In some embodiments, the UAV operates on undulating ground. For example, the UAV’s operating ground has appendages, such as trees, water towers, poles, or ponds, potholes, and the like. These terrains are not level enough, which will affect the normal operation of UAVs and even cause safety hazards. For example, the drone needs to maintain a safe height with the ground when operating on uneven ground, or the drone needs to land on a flat ground when landing.

In this scenario, if the ground flatness is predicted only based on the terrain data in front of the drone's nose, the resulting flatness cannot reflect the overall terrain environment where the drone is located. For example, the predicted flatness cannot reflect the unmanned In other positions of the drone, such as the terrain behind, on the left, and on the right, the safety of drone operations is poor.

Exemplarily, the terrain parameter of the omnidirectional scanning area determined in step S130 according to the fitting plane includes the flatness of the omnidirectional scanning area.

Exemplarily, the determining the terrain parameters of the omnidirectional scanning area according to the fitting plane includes: determining the coordinates of the multiple ground points to the fitting plane according to the coordinate data of the multiple ground points The average value of the distance is used to determine the flatness of the omnidirectional scanning area according to the average value.

Exemplarily, the flatness of the omnidirectional scanning area may be determined according to the average value of the distance from the filtered ground point, that is, the coordinate data of the effective ground point to the fitting plane.

Exemplarily, calculate the straight-line distance from the ground point {x _i , y _i , z _{i} to the fitting plane:}

Then calculate the average of the distances from ground points, for example, n ground points to the fitting plane:

For example, the average value of the distances from the n ground points to the fitting plane may be used as the flatness of the omnidirectional scanning area. If the average value is larger, it means that the ground in the omnidirectional scanning area is more uneven; if the average value is smaller, it means that the ground is flatter.

It is understandable that the flatness may also be determined by the residual of the distance from the ground point to the fitting plane, for example, the square of the residual of the distance from the n ground points to the fitting plane is determined as the flatness. However, the residual error cannot describe the flatness of the sloped ground well, and the existence of the slope will make the residual error value larger. Therefore, in order to eliminate the influence of the ground slope on the flatness, the ground flatness can be determined by using the mean value of the distance from the effective ground point to the fitting plane.

It is understandable that by acquiring the detection data of several ground points in the omnidirectional scanning area, and determining the fitting plane of the omnidirectional scanning area according to the detection data, the flatness of the omnidirectional scanning area can be determined according to the fitting plane. It can reflect the leveling of the terrain in different directions from the front, back, left, and right of the drone.

Exemplarily, the height value of the ground directly below the radar is determined according to the intercept of the fitting plane.

Exemplarily, the origin of the geodetic coordinate system is directly below the radar, and the intercept of the fitting plane may be equal to the height of the ground directly below the center of the radar.

In some embodiments, the adjusting the flight motion of the drone according to the terrain parameter in step S140 includes: adjusting at least one of a flight speed, a pitch angle, a roll angle, and a yaw angle according to the slope. .

For example, if the slope in the flying direction of the drone is large, you can reduce the flight speed and increase the pitch angle; if the slope in the left direction of the drone is small, you can fly to the left by adjusting the yaw angle .

Exemplarily, the projection of the flying speed direction of the drone on the fitting plane can be used to predict the terrain changes in the flying direction of the drone, and the speed of the drone can be controlled in advance, so that the drone can be used in mountains and other environments. Follow any direction, safely and quickly imitate the ground operation. It can solve the problem that the drone can only fly along the slope ascending/descending direction when operating in mountainous areas. It can detect any direction in real time, such as the topographical fluctuations relative to the current flying direction, so that the drone can travel along any direction in the mountains. Fly in the same direction.

Exemplarily, by predicting the terrain undulations in the flight direction, the speed control of the UAV can respond in advance, thereby ensuring the safety of the unmanned UAV operating in the mountains.

In some embodiments, adjusting the flight motion of the UAV according to the terrain parameters in step S140 includes: adjusting the flight speed, pitch angle, roll angle, and yaw according to the height value of the ground directly below the radar. At least one of the angles.

Exemplarily, in certain operating scenarios, such as spraying pesticides, the drone and the operating target need to be maintained. If the distance between the trees is kept within a range, the drone can be better controlled by determining the height of the ground directly below To maintain the distance between job targets.

For example, when the altitude value of the ground directly below the radar decreases, the descending height can be adjusted by adjusting the pitch angle; if the altitude value of the ground directly below the radar changes greatly, the drone can be controlled to reduce the flight speed to ensure flight safety.

In some embodiments, in step S140, adjusting the flight action of the drone according to the terrain parameter includes: adjusting at least one of a flight speed, a pitch angle, a roll angle, and a yaw angle according to the flatness. item.

For example, if the omnidirectional scanning area is very uneven, you can reduce the flight speed or reduce the pitch angle, roll angle, and yaw angle to ensure flight safety.

Exemplarily, the adjusting the flying action of the drone according to the terrain parameter includes: adjusting the speed control sense according to the flatness.

Sensitivity also becomes sensitivity. If the omnidirectional scanning area is very uneven, you can reduce the sensitivity of the speed control to have more time for obstacle avoidance and error correction.

In some embodiments, in step S140, adjusting the flight motion of the drone according to the terrain parameter includes: judging whether to land in the omnidirectional scanning area according to the flatness, and if so, in the omnidirectional scanning area. Landing towards the scanning area.

For example, if it is judged that the omnidirectional scanning area is sufficiently flat according to the flatness, it will land in the omnidirectional scanning area.

Since the flatness can reflect the terrain flatness of the UAV in different directions, the determined landing area is safer, and the autonomous landing point selection of the UAV based on the flatness of the ground can be realized.

In some embodiments, adjusting the flight action of the drone according to the terrain parameter in step S140 includes: determining a flight route according to the terrain parameter; and flying along the flight route.

Exemplarily, the flight route can be determined based on the slope of the omnidirectional scanning area in different directions, the flatness of the omnidirectional scanning area, the height of the ground and other terrain parameters, and/or the variation amplitude and frequency of the terrain parameters, etc., to achieve a safer Plan the flight route independently.

The drone operation control method provided by the embodiment of this specification acquires the detection data of the omnidirectional scanning area obtained by the radar carried by the drone during the ground rotation scanning process, and fits the omnidirectional scanning according to the detection data The fitting plane of the area; to realize the determination of the terrain parameters of the omnidirectional scanning area according to the fitting plane; and adjust the flight action of the drone according to the terrain parameters. Since the omnidirectional scanning area includes different directions of the UAV, the obtained terrain parameters are more global and accurate, so that the flight actions of the UAV can be controlled more safely.

Please refer to FIG. 12, which is a schematic block diagram of an unmanned aerial vehicle 700 according to an embodiment of the present specification. The UAV 700 includes an integrated radar, or may be additionally equipped with an independent radar.

The drone 700 includes a processor 701 and a memory 702.

Exemplarily, the processor 701 and the memory 702 are connected by a bus 703, and the bus 703 is, for example, an I2C (Inter-integrated Circuit) bus.

Specifically, the processor 701 may be a micro-controller unit (MCU), a central processing unit (CPU), a digital signal processor (Digital Signal Processor, DSP), or the like.

Specifically, the memory 702 may be a Flash chip, a read-only memory (ROM, Read-Only Memory) disk, an optical disk, a U disk, or a mobile hard disk.

Wherein, the processor 701 is configured to run a computer program stored in the memory 702, and implement the aforementioned drone operation control method when the computer program is executed.

Exemplarily, the processor 701 is configured to run a computer program stored in the memory 702, and implement the following steps when the computer program is executed:

Acquiring the omnidirectional scanning area detection data obtained by the radar during the ground rotation scanning process;

Fitting a fitting plane of the omnidirectional scanning area according to the detection data;

Determining the terrain parameter of the omnidirectional scanning area according to the fitting plane, the terrain parameter including the slope of the omnidirectional scanning area;

Adjust the flight action of the drone according to the terrain parameter.

The specific principles and implementation methods of the drone provided in the embodiment of this specification are similar to the drone operation control method of the foregoing embodiment, and will not be repeated here.

The embodiments of this specification also provide a computer-readable storage medium, the computer-readable storage medium stores a computer program, the computer program includes program instructions, and the processor executes the program instructions to implement the foregoing implementation The steps of the method provided by the example.

Wherein, the computer-readable storage medium may be the internal storage unit of the radar, the terrain prediction device, or the UAV described in any of the foregoing embodiments, for example, the hard disk or memory of the UAV. The computer-readable storage medium may also be an external storage device of the radar, a terrain prediction device, or a drone, such as a plug-in hard disk or a smart memory card (Smart Media Card, SMC) equipped on the drone, Secure Digital (SD) card, flash card (Flash Card), etc.

The UAV and the computer-readable storage medium provided in the above-mentioned embodiments of this specification acquire the detection data of the omnidirectional scanning area obtained by the radar carried by the UAV during the ground rotation scanning process, and fit the detection data according to the detection data. The fitting plane of the omnidirectional scanning area; to realize the determination of the terrain parameters of the omnidirectional scanning area according to the fitting plane; and adjusting the flight motion of the drone according to the terrain parameters. Since the omnidirectional scanning area includes the different directions of the UAV, the terrain parameters obtained are more global and accurate, so that the flight movements of the UAV can be controlled more safely.

It should be understood that the terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the specification.

It should also be understood that the term "and/or" used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

The above are only specific implementations of this specification, but the protection scope of this specification is not limited to this. Any person skilled in the art can easily think of various equivalents within the technical scope disclosed in this specification. Modifications or replacements, these modifications or replacements shall be covered within the protection scope of this manual. Therefore, the protection scope of this specification should be subject to the protection scope of the claims.

Claims (66)

1. A terrain prediction method for slopes, characterized in that the method includes:

   Obtain the detection data of the omnidirectional scanning area obtained by the radar during the ground rotation scanning process;

   Fitting a fitting plane of the omnidirectional scanning area according to the detection data;

   The terrain parameter of the omnidirectional scanning area is determined according to the fitting plane, and the terrain parameter includes the slope of the omnidirectional scanning area.
2. The method according to claim 1, wherein the terrain parameter of the omnidirectional scanning area further comprises at least one of the following:

   The flatness of the omnidirectional scanning area and the height value of the ground directly below the radar.
3. The method according to claim 1, wherein said acquiring the detection data of the omnidirectional scanning area obtained by the radar during the scanning of the ground rotation comprises:

   Obtain the detection range and azimuth angle of several ground points in the omnidirectional scanning area relative to the radar.
4. The method according to claim 3, wherein the method further comprises:

   Obtain the corresponding radar rotation angle when the radar detects each of the ground points.
5. The method according to claim 4, wherein said acquiring the detection data of the omnidirectional scanning area obtained by the radar during the scanning process of the ground rotation further comprises:

   The distance of each ground point relative to the radar in multiple directions is determined according to the radar rotation angle, detection distance, and azimuth angle corresponding to the ground point.
6. The method according to claim 1, wherein said acquiring the detection data of the omnidirectional scanning area obtained by the radar during the scanning of the ground rotation comprises:

   Obtain the distances of several ground points in the omnidirectional scanning area relative to the radar in multiple directions.
7. The method according to claim 6, wherein the distance of the ground point relative to the radar in multiple directions comprises:

   The position coordinates of the ground point in the radar coordinate system of the radar;

   The radar coordinate system takes the rotation center of the radar as the origin, the first axis direction is directly below the radar, the second axis direction is the front direction of the radar, and is perpendicular to the first axis. The direction between the direction and the second axis direction is the third axis direction.
8. The method according to any one of claims 1-7, wherein the fitting a fitting plane of the omnidirectional scanning area according to the detection data comprises:

   Determining, according to the detection data, coordinate data of several ground points in the omnidirectional scanning area in a geodetic coordinate system;

   Fitting the fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points.
9. The method according to claim 8, wherein the method further comprises:

   The attitude information of the radar is acquired through the inertial measurement unit carried by the drone and/or the inertial measurement unit carried by the radar.
10. The method according to claim 9, wherein the determining the coordinate data of each of the ground points in the omnidirectional scanning area in the geodetic coordinate system according to the detection data comprises:

    The coordinate data of each of the ground points in the geodetic coordinate system is determined according to the attitude information of the radar and the detection data.
11. The method according to claim 8, wherein the origin of the geodetic coordinate system is located directly below the radar, and the geodetic coordinate system uses the true north direction or the true south direction of the geodetic origin as the fourth axis direction, Taking the true east direction or the true west direction of the earth origin as the fifth axis direction, and taking the direction perpendicular to the fourth axis direction and the fifth axis direction as the sixth axis direction;

    The coordinate data of the ground point in the geodetic coordinate system includes the distance of the ground point in the fourth axis direction, the fifth axis direction, and the sixth axis direction relative to the geodetic origin.
12. The method according to claim 8, wherein the fitting a fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points comprises:

    The coordinate data of the several ground points are screened, and the fitting plane of the omnidirectional scanning area is fitted according to the screened coordinate data of the ground points.
13. The method according to claim 12, wherein the screening of the coordinate data of the several ground points, and fitting the fitting plane of the omnidirectional scanning area according to the coordinate data of the filtered ground points, include:

    Performing a cluster analysis on the plurality of ground points according to the coordinate data of the plurality of ground points, and determine the ground points that meet the clustering condition;

    Fitting the fitting plane of the omnidirectional scanning area according to the coordinate data of the ground points that meet the clustering condition.
14. The method according to claim 13, wherein the clustering analysis of the plurality of ground points according to the coordinate data of the plurality of ground points to determine the ground points that meet the clustering condition comprises:

    Cyclic execution: Determine one of the ground points as the points to be clustered, determine the number of ground points in the search range of the points to be clustered, and remove the points to be clustered if the number is not greater than the preset clustering threshold. Cluster points

    Until the several ground points are determined as points to be clustered;

    It is determined that the unremoved points to be clustered are ground points that meet the clustering condition.
15. The method according to claim 8, wherein the fitting a fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points comprises:

    Determining at least three ground points from the plurality of ground points, and determining a target plane according to the at least three ground points;

    Determining the plane distance from each of the ground points to the target plane according to the coordinate data of the several ground points;

    If the number of ground points whose plane distance is not greater than the distance threshold is not less than a preset number threshold, the fitting plane of the omnidirectional scanning area is obtained by fitting the ground points whose plane distance is not greater than the distance threshold.
16. The method according to claim 15, wherein the fitting a fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points further comprises:

    If the difference between the intercept of the fitted plane and the height values of all interior points is not less than the preset threshold value, return to the determination of at least three ground points from the plurality of ground points, and determine according to the at least three ground points A target plane; determine the plane distance from each of the ground points to the target plane according to the coordinate data of the several ground points; if the plane distance is not greater than the distance threshold, the number of ground points is not less than the preset number threshold, The step of obtaining the fitted plane of the omnidirectional scanning area by fitting the ground points whose plane distance is not greater than the distance threshold is continued, and the at least three ground points are different from at least one of the three ground points determined last time .
17. The method according to claim 8, wherein the determining the terrain parameters of the omnidirectional scanning area according to the fitting plane comprises:

    According to the coordinate data of the multiple ground points, an average value of the distances from the multiple ground points to the fitting plane is determined, and the flatness of the omnidirectional scanning area is determined according to the average value.
18. The method according to claim 2, wherein the determining the terrain parameters of the omnidirectional scanning area according to the fitting plane comprises:

    Determine the direction of the slope;

    Determine the slope of the fitting plane in the slope direction.
19. The method according to claim 18, wherein the determining the slope direction comprises:

    It is determined that the flying direction of the drone equipped with the radar is the gradient direction.
20. The method according to claim 18, wherein the determining the slope of the fitting plane in the slope direction comprises:

    The slope of the fitting plane in the slope direction is determined according to the normal vector of the fitting plane.
21. The method according to claim 2, wherein the determining the terrain parameters of the omnidirectional scanning area according to the fitting plane comprises:

    The height value of the ground directly below the radar is determined according to the intercept of the fitting plane.
22. The method according to any one of claims 1-7, wherein the radar is mounted on an unmanned aerial vehicle, the radar includes an antenna mechanism, and the antenna mechanism is capable of circling around the fuselage of the drone The preset rotation axis rotates to detect obstacles on the side of the drone; the radar is located under the bottom of the fuselage, and the rotation axis intersects a preset plane, which is the unmanned plane. The plane where the pitch axis and roll axis of the aircraft are located.
23. The method according to claim 22, wherein the angle between the rotation axis and the preset plane is 60°-90°.
24. The method according to claim 23, wherein the rotation axis is substantially perpendicular to the preset plane; and/or, the rotation axis is substantially coincident with the center line of the fuselage.
25. The method according to claim 23, wherein the radar is installed on the landing frame of the drone through a mounting structure, and the mounting structure is located between the radar and the fuselage.
26. The method according to claim 22, wherein the rotation axis and the yaw axis of the drone are at an acute angle.
27. The method according to claim 26, wherein the acute angle is 0°-30°.
28. The method according to claim 22, wherein the rotation angle range of the antenna mechanism around the rotating shaft is greater than or equal to 360°.
29. The method according to any one of claims 1-7, wherein the radar is a continuous wave radar.
30. A terrain prediction device for slopes, characterized in that the terrain prediction device includes a memory and a processor;

    The memory is used to store a computer program;

    The processor is configured to execute the computer program and, when executing the computer program, implement the following steps:

    Obtain the detection data of the omnidirectional scanning area obtained by the radar during the ground rotation scanning process;

    Fitting a fitting plane of the omnidirectional scanning area according to the detection data;

    The terrain parameter of the omnidirectional scanning area is determined according to the fitting plane, and the terrain parameter includes the slope of the omnidirectional scanning area.
31. A radar, characterized in that the radar includes a memory and a processor;

    The memory is used to store a computer program;

    The processor is configured to execute the computer program and, when executing the computer program, implement the following steps:

    Obtain the detection data of the omnidirectional scanning area obtained by the radar during the ground rotation scanning process;

    Fitting a fitting plane of the omnidirectional scanning area according to the detection data;

    The terrain parameter of the omnidirectional scanning area is determined according to the fitting plane, and the terrain parameter includes the slope of the omnidirectional scanning area.
32. An unmanned aerial vehicle operation control method, the unmanned aerial vehicle is equipped with radar and operates on slopes, characterized in that the method includes:

    Acquiring the omnidirectional scanning area detection data obtained by the radar during the ground rotation scanning process;

    Fitting a fitting plane of the omnidirectional scanning area according to the detection data;

    Determining the terrain parameter of the omnidirectional scanning area according to the fitting plane, the terrain parameter including the slope of the omnidirectional scanning area;

    Adjust the flight action of the drone according to the terrain parameter.
33. The method according to claim 32, wherein the terrain parameter of the omnidirectional scanning area further comprises at least one of the following:

    The flatness of the omnidirectional scanning area and the height value of the ground directly below the radar.
34. The method according to claim 32, wherein said acquiring the detection data of the omnidirectional scanning area obtained by the radar during the ground rotation scanning process comprises:

    Obtain the detection range and azimuth angle of several ground points in the omnidirectional scanning area relative to the radar.
35. The method according to claim 34, wherein the method further comprises:

    Obtain the corresponding radar rotation angle when the radar detects each of the ground points.
36. The method according to claim 35, wherein said acquiring the detection data of the omnidirectional scanning area obtained by the radar during the ground rotation scanning process further comprises:

    The distance of each ground point relative to the radar in multiple directions is determined according to the radar rotation angle, detection distance, and azimuth angle corresponding to the ground point.
37. The method according to claim 32, wherein said acquiring the detection data of the omnidirectional scanning area obtained by the radar during the ground rotation scanning process comprises:

    Obtain the distances of several ground points in the omnidirectional scanning area relative to the radar in multiple directions.
38. The method according to claim 37, wherein the distance of the ground point relative to the radar in multiple directions comprises:

    The position coordinates of the ground point in the radar coordinate system of the radar;

    The radar coordinate system takes the rotation center of the radar as the origin, the first axis direction is directly below the radar, the second axis direction is the front direction of the radar, and is perpendicular to the first axis. The direction between the direction and the second axis direction is the third axis direction.
39. The method according to any one of claims 32-38, wherein the fitting a fitting plane of the omnidirectional scanning area according to the detection data comprises:

    Determining, according to the detection data, coordinate data of several ground points in the omnidirectional scanning area in a geodetic coordinate system;

    Fitting the fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points.
40. The method according to claim 39, wherein the method further comprises:

    The attitude information of the radar is acquired through the inertial measurement unit carried by the unmanned aerial vehicle and/or the inertial measurement unit carried by the radar.
41. The method according to claim 40, wherein the determining, according to the detection data, the coordinate data of several ground points in the omnidirectional scanning area in a geodetic coordinate system comprises:

    The coordinate data of several ground points in the geodetic coordinate system is determined according to the attitude information of the radar and the detection data.
42. The method according to claim 39, wherein the origin of the geodetic coordinate system is located directly below the radar, and the geodetic coordinate system takes the true north or true south direction of the geodetic origin as the fourth axis direction, Taking the true east direction or the true west direction of the earth origin as the fifth axis direction, and taking the direction perpendicular to the fourth axis direction and the fifth axis direction as the sixth axis direction;

    The coordinate data of the ground point in the geodetic coordinate system includes the distance of the ground point in the fourth axis direction, the fifth axis direction, and the sixth axis direction relative to the geodetic origin.
43. The method according to claim 39, wherein the fitting a fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points comprises:

    The coordinate data of the several ground points are screened, and the fitting plane of the omnidirectional scanning area is fitted according to the screened coordinate data of the ground points.
44. The method according to claim 43, wherein the screening of the coordinate data of the several ground points, and fitting the fitting plane of the omnidirectional scanning area according to the coordinate data of the filtered ground points, include:

    Performing a cluster analysis on the plurality of ground points according to the coordinate data of the plurality of ground points, and determine the ground points that meet the clustering condition;

    Fitting the fitting plane of the omnidirectional scanning area according to the coordinate data of the ground points that meet the clustering condition.
45. The method according to claim 44, wherein the performing cluster analysis on the plurality of ground points according to the coordinate data of the plurality of ground points to determine the ground points that meet the clustering condition comprises:

    Cyclic execution: Determine one of the ground points as the points to be clustered, determine the number of ground points in the search range of the points to be clustered, and remove the points to be clustered if the number is not greater than the preset clustering threshold. Cluster points

    Until the several ground points are determined as points to be clustered;

    It is determined that the unremoved points to be clustered are ground points that meet the clustering condition.
46. The method according to claim 39, wherein the fitting a fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points comprises:

    Determining at least three ground points from the plurality of ground points, and determining a target plane according to the at least three ground points;

    Determining the plane distance from each of the ground points to the target plane according to the coordinate data of the several ground points;

    If the number of ground points whose plane distance is not greater than the distance threshold is not less than a preset number threshold, the fitting plane of the omnidirectional scanning area is obtained by fitting the ground points whose plane distance is not greater than the distance threshold.
47. The method according to claim 46, wherein the fitting of the fitting plane of the omnidirectional scanning area according to the coordinate data of the several ground points further comprises:

    If the difference between the intercept of the fitted plane and the height values of all interior points is not less than the preset threshold value, return to the determination of at least three ground points from the plurality of ground points, and determine according to the at least three ground points A target plane; determine the plane distance from each of the ground points to the target plane according to the coordinate data of the several ground points; if the plane distance is not greater than the distance threshold, the number of ground points is not less than the preset number threshold, The step of obtaining the fitted plane of the omnidirectional scanning area by fitting the ground points whose plane distance is not greater than the distance threshold is continued, and the at least three ground points are different from at least one of the three ground points determined last time .
48. The method according to claim 39, wherein the determining the terrain parameters of the omnidirectional scanning area according to the fitting plane comprises:

    According to the coordinate data of the multiple ground points, an average value of the distances from the multiple ground points to the fitting plane is determined, and the flatness of the omnidirectional scanning area is determined according to the average value.
49. The method according to any one of claims 32-38, wherein the determining the terrain parameters of the omnidirectional scanning area according to the fitting plane comprises:

    Determine the direction of the slope;

    Determine the slope of the fitting plane in the slope direction.
50. The method according to claim 49, wherein the determining the slope direction comprises:

    It is determined that the flying direction of the drone is the gradient direction.
51. The method according to claim 49, wherein the determining the slope of the fitting plane in the slope direction comprises:

    The slope of the fitting plane in the slope direction is determined according to the normal vector of the fitting plane.
52. The method according to claim 33, wherein the determining the terrain parameters of the omnidirectional scanning area according to the fitting plane comprises:

    The height value of the ground directly below the radar is determined according to the intercept of the fitting plane.
53. The method according to any one of claims 32-38, wherein the adjusting the flight action of the drone according to the terrain parameter comprises:

    Adjust at least one of flight speed, pitch angle, roll angle, and yaw angle according to the slope.
54. The method according to claim 33 or 52, wherein the adjusting the flight action of the drone according to the terrain parameter comprises:

    Adjust at least one of the flight speed, pitch angle, roll angle, and yaw angle according to the altitude value of the ground directly below the radar.
55. The method according to claim 39, wherein the adjusting the flight action of the drone according to the terrain parameter comprises:

    Adjust at least one of the flight speed, pitch angle, roll angle, and yaw angle according to the flatness of the omnidirectional scanning area; and/or

    Adjust the sense of speed control according to the flatness; and/or

    Determine whether to land in the omnidirectional scan area according to the flatness, and if so, land in the omnidirectional scan area.
56. The method according to claim 33 or 52, wherein the adjusting the flight action of the drone according to the terrain parameter comprises:

    Determine the flight route according to the terrain parameters;

    Fly along the flight path.
57. The method according to any one of claims 32-38, wherein the radar is mounted on an unmanned aerial vehicle, the radar includes an antenna mechanism, and the antenna mechanism is capable of circling around the fuselage of the unmanned aerial vehicle. The preset rotation axis rotates to detect obstacles on the side of the drone; the radar is located under the bottom of the fuselage, and the rotation axis intersects a preset plane, which is the unmanned plane. The plane where the pitch axis and roll axis of the aircraft are located.
58. The method according to claim 57, wherein the angle between the rotation axis and the preset plane is 60°-90°.
59. The method according to claim 58, wherein the rotation axis is substantially perpendicular to the preset plane; and/or, the rotation axis is substantially coincident with the center line of the fuselage.
60. The method according to claim 58, wherein the radar is installed on the landing frame of the drone through a mounting structure, and the mounting structure is located between the radar and the fuselage.
61. The method according to claim 57, wherein the rotation axis and the yaw axis of the drone are at an acute angle.
62. The method according to claim 61, wherein the acute angle is 0°-30°.
63. The method according to claim 57, wherein the rotation angle range of the antenna mechanism around the rotating shaft is greater than or equal to 360°.
64. The method according to any one of claims 32-38, wherein the radar is a continuous wave radar.
65. An unmanned aerial vehicle equipped with radar, characterized in that the unmanned aerial vehicle includes a memory and a processor;

    The memory is used to store a computer program;

    The processor is configured to execute the computer program and, when executing the computer program, implement the following steps:

    Acquiring the omnidirectional scanning area detection data obtained by the radar during the ground rotation scanning process;

    Fitting a fitting plane of the omnidirectional scanning area according to the detection data;

    Determining the terrain parameter of the omnidirectional scanning area according to the fitting plane, the terrain parameter including the slope of the omnidirectional scanning area;

    Adjust the flight action of the drone according to the terrain parameter.
66. A computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor realizes:

    The method of any one of claims 1-29; and/or

    The method of any one of claims 32-64.

Images