CN104102218B - The perception of view-based access control model servo and bypassing method and system - Google Patents

Abstract

The present invention provides a visual servo-based perception and avoidance system, including a drone, an image acquisition system, a visual servo control system, and a navigation and positioning system. The visual servo system includes a visual target detection and tracking module, a security envelope module and Visual Servo Controller. The present invention also provides an avoidance method based on the sense-and-avoid system. The present invention realizes rapid perception of the surrounding airspace environment, and does not require the intervention and operation of ground operators without installing any ranging sensors. Under the environment, the evasive maneuver of the space flight target can be independently completed; the invention has low load requirements, high control precision, high intelligence, and can improve the airspace flight safety capability of the UAV.

Description

Perception and avoidance method and system based on visual servoing

technical field

The invention relates to the field of unmanned aerial vehicle navigation and control, in particular to a visual servo-based perception and avoidance method and system.

Background technique

Recently, with the increasingly strong demand for UAVs in military applications and civilian fields, coupled with the further opening of my country's civil airspace, the future airspace will present a situation where various types of functional UAVs and manned aircraft will share and integrate airspace , the airspace will become increasingly dense. In this case, the UAV's Sense and Avoid (SAA) capability will become a prerequisite for entering the airspace and ensuring the safety of UAV flight. UAV perception and avoidance means that UAVs use airborne sensors or ground surveillance systems to monitor the airspace flight environment and obtain the status of flying targets, plan evasive paths for targets with potential collision threats, and complete evasive maneuvers. Ensure the flight safety of drones.

SAA technology is a key technical problem to be solved urgently in the field of UAV technology. In 2013, in the UAV system airspace integration roadmap released by the US FAA (Federal Aviation Administration), it was clearly stated that SAA capability is a necessary capability for UAVs to fly in national airspace. Its main functions are divided into: target detection and tracking, collision threat estimation, evasive path planning, and evasive maneuvering.

UAV airborne SAA sensors include T-CAS, ADS-B, radar, lidar, photoelectric, infrared sensors, etc. Among them, photoelectric sensors are very important and effective means of perception and avoidance for non-cooperative targets, especially in the case of light and small UAV systems with limited load, tasks, and cost, and cannot carry large-scale sensing equipment such as radar and lidar. The advantages of the sensor's visual SAA system in terms of quality, power consumption, and cost make it easier to integrate and apply in light and small systems; the characteristics of passive perception of visual SAA are suitable for specific tasks such as battlefield environments and stealth tasks; and ADS-B, Compared with TCAS, etc., visual SAA does not rely on the response mechanism, does not need ground control information and satellite positioning information, that is, it can provide independent perception and avoidance capabilities for non-cooperative targets. Therefore, it is of great significance to study the visual SAA system based on photoelectric sensor.

Contents of the invention

The purpose of the present invention is to provide a visual servo-based perception and avoidance method and system to solve the problems of existing sensors with large volume and high power consumption, which are not conducive to the integrated application in light and small unmanned aerial vehicles systems, and the present invention The invention can realize the rapid perception of the surrounding airspace environment, establish a visual feedback control model through the angle safety envelope, and control the UAV to complete the perception and avoidance of airspace targets.

The present invention provides a perception and avoidance system based on visual servoing, including:

An unmanned aerial vehicle, an autopilot is set on the unmanned aerial vehicle, and the automatic pilot receives flight control instructions to fly and complete the avoidance action of the unmanned aerial vehicle to obstacles;

The image acquisition system is mounted on the unmanned aerial vehicle and is used to acquire the flight airspace image information of the unmanned aerial vehicle, adopts a multi-camera multi-view centralized array configuration, and transmits image information in real time through the camera interface;

The visual servo control system is arranged on the unmanned aerial vehicle and includes a visual target detection and tracking module, a safety envelope module and a visual servo controller. The visual target detection and tracking module receives the image information transmitted by the image acquisition system and realizes the detection For the detection, positioning and tracking of flying targets, the safety envelope module generates the image plane angle safety envelope based on the acquired target state information, and generates image avoidance angles based on the safety envelope, and inputs this angle to the visual servo controller. The visual servo controller outputs the evasive maneuvering angle, and transmits the angle to the autopilot through the serial port to complete the evasive action;

The navigation and positioning system is set on the UAV and is used to establish a route return mechanism. When the navigation and positioning system detects that the UAV is not on the route and there is no threat, it performs route regression to make the UAV return to the original route.

As a preferred mode, the visual target detection and tracking module in the present invention adopts the multi-channel high-definition image processing platform DM8168.

The present invention also provides a method for avoiding the perception and avoidance system based on visual servoing, including:

Step 1, realize the complete coverage of the airspace within the flight field of view through the UAV airborne image acquisition system, and obtain the flight airspace image information of the UAV;

Step 2, the visual target detection and tracking module in the airborne visual servo control system receives and processes the image information in step 1, and completes the detection, positioning and tracking of the flying target in the airspace image;

Step 3: According to the state information of the flight target obtained in step 2, the safety envelope module in the visual servo control system generates the image plane angle safety envelope, and determines whether there is a collision threat to the UAV through the safety envelope. For the collision threat, calculate Obtain the avoidance angle in the image sense;

Step 4, input the evasive angle in the step 3 into the visual servo in the visual servo control system, the visual servo outputs the required evasive maneuver pitch and yaw angle, and transmits to the UAV autopilot to complete the evasive maneuver;

Step five, use the navigation and positioning system to establish a route return mechanism, and when the UAV is not threatened, the UAV will return to the original route.

As a preferred method, the specific implementation method of the second step is as follows:

In step 1, the image data acquired by the image acquisition system is transmitted to the visual target detection and tracking module in the airborne visual servo control system through the camera data cable. The image acquisition system is equipped with three cameras A, B, and C for visual target detection and tracking. The module adopts the multi-channel high-definition image processing platform DM8168, and uses the DM8168 to process the acquired multi-channel video, and obtains the position information p k of a certain camera image plane in the image acquisition system at time k of the target through the detection algorithm p _k =[p _xk ,p _yk ] to get the relative angle of the target relative to the camera:

Among them, w and h are the width and height of the image in pixels, respectively, f is the focal length of the camera lens, and μ is the pixel size;

According to the installation position of the camera, the relative angle of the target relative to the UAV at moment k can be obtained as Θ _k = [σ _k ,γ _k ], where

γ _k = γ′

In the formula Respectively indicate that the detected target comes from camera A, camera B and camera C in the image acquisition system;

Determine the angular velocity of the target Ω _k =[ω _xk ,ω _yk ]=Θ _k+1 -Θ _k ,k>0 according to the angular changes at the previous and subsequent moments.

As another preferred manner, the step of generating the safety envelope and the step of calculating the avoidance angle in the step 3 include:

①Establish the target safety enveloping circle, the center of which is the target point, select an initial value of the radius, and grow at a fixed speed;

②Create a safe envelope ellipse according to the size of the target. The center of the ellipse is the target point. The length of the minor axis of the ellipse is equal to the radius of the envelope circle. The sum of the extrapolated lengths of , the long axis direction is parallel to the moving direction of the target;

③ The aforementioned generated envelope circle and envelope ellipse form a plane angle safety envelope, which is composed of a semi-ellipse along the direction of target motion and a semi-circle along the opposite direction of target motion with the short axis of the ellipse as the dividing line;

④ Estimate the target threat based on the aforementioned plane angle safety envelope, and establish a threat marking function based on the safety envelope to judge whether there is a collision threat;

⑤ According to the principle of minimum avoidance distance, find the point closest to the initial target in the plane angle safety envelope.

In particular, the specific generation steps of the safety envelope and the specific calculation steps of the avoidance angle in the step three are as follows:

①Establish the target safety enveloping circle, select the center of the circle as the relative angular position Θ _k =[σ _k , γ _k ] of the target relative to the UAV at time k, select an initial value r ₀ of the radius, and increase it at a fixed speed, then

r _k+1 =r _k +ε,k=1,2,3...

r _k = r ₀ , k = 0

Where ε is a fixed growth rate, r _k+1 represents the radius of the safe envelope circle at time k+1;

② Establish a safety envelope ellipse according to the size of the target, the center of the ellipse is the target point Θ _k = [σ _k , γ _k ], the long axis direction is parallel to the target’s movement direction, that is, the direction of the target angular velocity Ω _k , the long axis and the short axis The length is selected according to the following formula:

a=r _k (1+α‖Ω _k ‖),

b=r _k ,

In the formula, a and b represent the major axis and minor axis of the envelope ellipse respectively, r _k is the radius of the envelope circle at time k, α is the scaling factor for converting angular velocity into axis length, which represents the reserved time for avoidance, ‖Ω _k ‖ is the modulus of the target motion angular velocity vector;

③ The aforementioned generated envelope circle and envelope ellipse form a plane angle safety envelope, which is composed of a semi-ellipse along the direction of target motion and a semi-circle along the opposite direction of target motion with the short axis of the ellipse as the dividing line;

④ Estimate the target threat based on the above-mentioned plane angle safety envelope, and establish a threat marking function l _k based on time k. When l _k <0, it is considered that there is a collision threat, so as to judge whether there is a collision threat, where

in Indicates the angle of the main axis direction of camera B relative to the main axis direction of the UAV, which can usually be regarded as is a sign function;

⑤When l _k <0, according to the principle of the minimum avoidance distance, find the distance in the plane angle safety envelope The nearest point s ^* = [σ ^* ,γ ^* ] as the minimum avoidance maneuver point:

in Represents the unit vector perpendicular to the target angular velocity vector at time k.

As yet another preferred manner, the specific calculation steps of avoiding maneuver pitch and yaw angles in the step 4 are as follows:

① Define the interaction matrix as

The speed controller is established in the visual servo controller of the visual servo control system, and the feedback amount is The exponential attenuation of the error is completed, and the output speed control value V is expressed as

In the formula, λ is the error exponential decay coefficient, Represents the pseudo-inverse of L _e ;

Since the information of the distance d between the target and the machine cannot be obtained in the interaction matrix L _e , the interaction matrix is divided as follows:

In the formula, L _ω represents the angle control component in L _e , and L _t represents the speed motion control component;

Get a distance-independent visual servo controller:

In the formula, represents the pseudo-inverse of L _ω ;

In order to eliminate the steady-state error, the visual servo controller is compensated as follows:

In the formula, Ω represents the angular velocity of the target movement, Indicates the direction vector of the error, ω _x , ω _y , ω _z represent the three components of ω respectively;

②The output pitch angle and yaw angle are

In the formula, and φ _c (k) represent the pitch and yaw angles of the autopilot output attitude, and φ(k) represent the pitch angle and yaw angle of the current attitude of the UAV, Δt represents the output time interval step of the visual servo controller, ω _y (k) and ω _z (k) represent the visual servo controller at the current moment The output pitch rate control amount and yaw rate control amount, g represents the acceleration of gravity, and V(k) represents the speed of the drone at the current moment.

As another preferred mode, the specific steps of establishing the route return mechanism in the step five are as follows:

①Use the navigation and positioning system of the aircraft to determine whether the UAV is on the route, and return to the route when the aircraft is not on the route;

②Confirm the return route angle at time k

In the formula, Represents the position of the minimum maneuver avoidance point when the route returns;

Where P _d is the position of the own aircraft on the original route when there is no evasive maneuver, P _d12 and P _d13 are the two-dimensional positions of the own aircraft on the XY and XZ planes respectively;

P _s is the current position of the machine obtained through the local navigation and positioning system, and P _s12 and P _s13 are the two-dimensional positions of the machine on the XY and XZ planes respectively;

e _s is the local direction vector obtained by the local navigation and positioning system, and e _s12 and e _s13 are the two-dimensional direction vectors of the local plane on the XY and XZ planes, respectively.

The invention realizes the rapid perception of the surrounding airspace environment, and independently completes the evasive maneuver of the space flight target without installing any ranging sensor and without the intervention and operation of ground operators; the invention has low load requirements , high control precision, and high intelligence, which can improve the airspace flight safety capability of UAVs.

Description of drawings

Fig. 1 is a structural block diagram of the perception and avoidance system based on visual servoing of the present invention;

FIG. 2 is a flow chart of the avoidance method of the visual servo-based perception and avoidance system of the present invention;

Figure 3a is a layout diagram of the camera array of the image acquisition system of the present invention;

Fig. 3b is a schematic diagram of the field of view angle of the camera array of the image acquisition system of the present invention;

FIG. 4 is a schematic diagram of an image plane angle security envelope of the present invention.

detailed description

The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

As shown in Figure 1, the present invention provides a perception and avoidance system based on visual servoing, including:

An unmanned aerial vehicle, an autopilot is set on the unmanned aerial vehicle, and the automatic pilot receives flight control instructions to fly and complete the avoidance action of the unmanned aerial vehicle to obstacles;

The image acquisition system is mounted on the unmanned aerial vehicle and is used to acquire the flight airspace image information of the unmanned aerial vehicle, adopts a multi-camera multi-view centralized array configuration, and transmits image information in real time through the camera interface;

The visual servo control system is arranged on the unmanned aerial vehicle and includes a visual target detection and tracking module, a safety envelope module and a visual servo controller. The visual target detection and tracking module receives the image information transmitted by the image acquisition system and realizes the detection For the detection, positioning and tracking of flying targets, the safety envelope module generates the image plane angle safety envelope based on the acquired target state information, and generates image avoidance angles based on the safety envelope, and inputs this angle to the visual servo controller. The visual servo controller outputs the evasive maneuvering angle, and transmits the angle to the autopilot through the serial port to complete the evasive action;

The navigation and positioning system is set on the UAV and is used to establish a route return mechanism. When the navigation and positioning system detects that the UAV is not on the route and there is no threat, it performs route regression to make the UAV return to the original route.

Among them, the visual target detection and tracking module in the present invention adopts the multi-channel high-definition image processing platform DM8168.

As shown in Figure 2, the present invention also provides a method for avoiding the perception and avoidance system based on visual servoing, including:

Step 1: Complete coverage of the airspace within the flight field of view is achieved through the UAV airborne image acquisition system, and image information of the UAV's flight airspace is acquired.

According to the regulations of the International Civil Aviation Organization, the field of view of the aircraft space must meet the following requirements: horizontal field of view>±110°, vertical field of view>±15°, so the above-mentioned airspace must be completely covered, and it should be as clear and accurate as possible Information. The image acquisition system adopts a multi-camera multi-view centralized array configuration. From the camera point of view, it must meet the needs of a large market and long-distance high-resolution imaging. However, it is difficult for a single camera lens to balance the field of view and focal length. Therefore, a multi-camera centralized array configuration as shown in Figure 3a is adopted. A, B, and C in the figure are all cameras set on the stabilized gimbal. The stabilized gimbal can ensure the acquisition of clear, horizontal perspective images. Using the focal length stitching of multiple telephoto lenses to obtain a large field of view angle, the specific field of view angle is shown in Figure 3b, and at the same time improve the detection ability of long-distance targets.

In step two, the visual target detection and tracking module in the airborne visual servo control system receives and processes the image information in step one, and completes the detection, positioning and tracking of the flying target in the airspace image.

In light and small UAVs, due to the limitation of communication bandwidth, it is often impossible to realize real-time transmission of high-definition images, and the long-distance transmission of multiple high-definition videos is difficult to achieve in the current light and small UAV systems.

The image data acquired by the image acquisition system is transmitted to the multi-channel high-definition image processing platform DM8168 in the airborne visual servo control system through the camera data line, and the DM8168 is used to process the acquired multi-channel video, and the target k time is acquired through the detection algorithm. The position information of a certain camera image plane in the system p _k =[p _xk ,p _yk ], to obtain the relative angle of the target relative to the camera:

Among them, w and h are the width and height of the image in pixels, respectively, f is the focal length of the camera lens, and μ is the pixel size;

According to the installation position of the camera, the relative angle of the target relative to the UAV at moment k can be obtained as Θ _k = [σ _k ,γ _k ], where

γ _k = γ′

In the formula Respectively indicate that the detected targets come from camera A, camera B and camera C in the image acquisition system.

Determine the angular velocity of the target Ω _k =[ω _xk ,ω _yk ]=Θ _k+1 -Θ _k ,k>0 according to the angular changes at the previous and subsequent moments.

Step 3: According to the state information of the flight target obtained in step 2, the safety envelope module in the visual servo control system generates the image plane angle safety envelope, and determines whether there is a collision threat to the UAV through the safety envelope. For the collision threat, calculate The avoidance angle s ^* = [σ ^* ,γ ^* ] in the image sense is obtained.

The safety envelope is defined as a smooth, closed curve or surface that represents the minimum separation distance between the aircraft and the target. In the visual avoidance system, since the distance cannot be obtained, the safety envelope cannot be established according to the real three-dimensional space of the target, as shown in Figure 4. Set up a kind of security envelope method based on angle among the present invention, comprise the steps:

①Establish the target safety enveloping circle, the center of which is the target point, select an initial value of the radius, and grow at a fixed speed;

②Create a safe envelope ellipse according to the size of the target. The center of the ellipse is the target point. The length of the minor axis of the ellipse is equal to the radius of the envelope circle. The sum of the extrapolated lengths of , the long axis direction is parallel to the moving direction of the target;

③ The aforementioned generated envelope circle and envelope ellipse form a plane angle safety envelope, which is composed of a semi-ellipse along the direction of target motion and a semi-circle along the opposite direction of target motion with the short axis of the ellipse as the dividing line;

④ Estimate the target threat based on the aforementioned plane angle safety envelope, and establish a threat marking function based on the safety envelope to judge whether there is a collision threat;

⑤ According to the principle of minimum avoidance distance, find the point closest to the initial target in the plane angle safety envelope.

The above method for establishing an angle-based security envelope can be carried out according to the following specific steps:

①Establish the target safety enveloping circle, select the center of the circle as the relative angular position Θ _k ＝[σ _k ,γ _k ] of the target relative to the UAV at moment k, select an initial value r ₀ of the radius, and increase it at a fixed speed, then

r _k+1 =r _k +ε,k=1,2,3...

r _k = r ₀ , k = 0

Where ε is a fixed growth rate, which is used to simulate the expansion speed of a three-dimensional space object on the image plane, r _k+1 represents the radius of the safe envelope circle at the moment k+1, and r in Figure 4 represents the radius of the envelope circle;

② Establish a safety envelope ellipse according to the size of the target. The center of the ellipse is the target point Θ _k = [σ _k , γ _k ], and the major axis direction is parallel to the moving direction of the target, which is the direction of the target angular velocity Ω _k . b represent the major axis and the minor axis of the envelope ellipse respectively, and the lengths of the major axis and the minor axis are selected according to the following formulas:

a=r _k (1+α‖Ω _k ‖),

b=r _k ,

In the formula, a and b represent the major axis and minor axis of the envelope ellipse respectively, r _k is the radius of the envelope circle at time k, α is the scaling factor for converting angular velocity into axis length, which represents the reserved time for avoidance, ‖Ω _k ‖ is the modulus of the target motion angular velocity vector, that is, the length of the major axis is the sum of the length of the minor axis and the extrapolated length of the target within the reserved time under the current angular velocity;

③ The envelope circle and envelope ellipse generated above form a plane angle safety envelope, which is composed of a semi-ellipse along the target movement direction and a semi-circle along the opposite direction of the target movement with the short axis of the ellipse as the dividing line, specifically as follows Shown in the solid line part in Fig. 4;

The safety envelope not only considers the avoidance angle margin due to the target approach and joint measurement error, but also considers the target's motion state;

④ Estimate the target threat based on the above-mentioned plane angle safety envelope, and establish a threat marking function l _k based on time k. When l _k <0, it is considered that there is a collision threat, so as to judge whether there is a collision threat, where

in Indicates the angle of the main axis direction of camera B relative to the main axis direction of the UAV, which can usually be regarded as is a sign function;

⑤When l _k <0, according to the principle of the minimum avoidance distance, find the distance in the plane angle safety envelope The nearest point s ^* = [σ ^* ,γ ^* ] as the minimum avoidance maneuver point:

in Represents the unit vector perpendicular to the target angular velocity vector at time k.

Step 4, input the avoidance angle s ^* = [σ ^* , γ ^* ] in step 3 into the visual servo in the visual servo control system, and the visual servo outputs the required evasive maneuver pitch and yaw angles And send it to the UAV autopilot through the serial port to complete the evasive maneuver.

① Define the interaction matrix as

The speed controller is established in the visual servo controller of the visual servo control system, and the feedback amount is The exponential attenuation of the error is completed, and the output speed control value V is expressed as

In the formula, λ is the error exponential decay coefficient, Represents the pseudo-inverse of L _e ;

Since the information of the distance d between the target and the machine cannot be obtained in the interaction matrix L _e , the interaction matrix is divided as follows:

In the formula, L _ω represents the angle control component in L _e , and L _t represents the speed motion control component;

Get a distance-independent visual servo controller:

In the formula, represents the pseudo-inverse of L _ω ;

In order to eliminate the steady-state error, the visual servo controller is compensated as follows:

In the formula, Ω represents the angular velocity of the target movement, Indicates the direction vector of the error, ω _x , ω _y , ω _z represent the three components of ω respectively;

②The output pitch angle and yaw angle are

In the formula, and φ _c (k) represent the pitch and yaw angles of the autopilot output attitude, and φ(k) represent the pitch angle and yaw angle of the current attitude of the UAV, Δt represents the output time interval step of the visual servo controller, ω _y (k) and ω _z (k) represent the visual servo controller at the current moment The output pitch rate control amount and yaw rate control amount, g represents the acceleration of gravity, and V(k) represents the speed of the drone at the current moment.

Step five, use the navigation and positioning system to establish a route return mechanism, and when the UAV is not threatened, the UAV will return to the original route.

①Use the navigation and positioning system of the aircraft to determine whether the UAV is on the route, and return to the route when the aircraft is not on the route;

②Confirm the return route angle at time k

In the formula, Represents the position of the minimum maneuver avoidance point when the route returns;

Where P _d is the position of the own aircraft on the original route when there is no evasive maneuver, P _d12 and P _d13 are the two-dimensional positions of the own aircraft on the XY and XZ planes respectively;

P _s is the current position of the machine obtained through the local navigation and positioning system, and P _s12 and P _s13 are the two-dimensional positions of the machine on the XY and XZ planes respectively;

e _s is the local direction vector obtained by the local navigation and positioning system, and e _s12 and e _s13 are the two-dimensional direction vectors of the local plane on the XY and XZ planes, respectively.

The method and system provided by the present invention can realize the evasive maneuver of the space flight target without installing any ranging sensor and without the intervention of the operator, with low load requirements, high control precision, and high Intelligence can improve the airspace flight safety capability of drones.

Claims (7)

1. A perception and avoidance system based on visual servoing, characterized in that it comprises: An unmanned aerial vehicle, an autopilot is set on the unmanned aerial vehicle, and the automatic pilot receives flight control instructions to fly and complete the avoidance action of the unmanned aerial vehicle to obstacles; The image acquisition system is mounted on the unmanned aerial vehicle and is used to acquire the flight airspace image information of the unmanned aerial vehicle, adopts a multi-camera multi-view centralized array configuration, and transmits image information in real time through the camera interface; The visual servo control system is arranged on the unmanned aerial vehicle and includes a visual target detection and tracking module, a safety envelope module and a visual servo controller. The visual target detection and tracking module receives the image information transmitted by the image acquisition system and realizes the detection For the detection, positioning and tracking of flying targets, the safety envelope module generates the image plane angle safety envelope based on the acquired target state information, and generates image avoidance angles based on the safety envelope, and inputs this angle to the visual servo controller. The visual servo controller outputs the evasive maneuvering angle, and transmits the angle to the autopilot through the serial port to complete the evasive action; The navigation and positioning system is set on the UAV and is used to establish a route return mechanism. When the navigation and positioning system detects that the UAV is not on the route and there is no threat, it performs route regression to make the UAV return to the original route. 2. The perception and avoidance system based on visual servoing according to claim 1, wherein the visual target detection and tracking module adopts a multi-channel high-definition image processing platform DM8168. 3. A method for avoiding the perception and avoidance system based on visual servoing, characterized in that it comprises: Step 1, realize the complete coverage of the airspace within the flight field of view through the UAV airborne image acquisition system, and obtain the flight airspace image information of the UAV; Step 2, the visual target detection and tracking module in the airborne visual servo control system receives and processes the image information in the step 1, and completes the detection, positioning and tracking of the flying target in the airspace image; The concrete realization method of described step 2 is as follows: The image data acquired by the image acquisition system in the step 1 is transmitted to the visual target detection and tracking module in the airborne visual servo control system through the camera data line, wherein the image acquisition system is equipped with three cameras A, B, and C, and the visual target detection The tracking and tracking module adopts the multi-channel high-definition image processing platform DM8168, utilizes the DM8168 to realize the processing of the acquired multi-channel video, and obtains the position information p k of a certain camera image plane in the image acquisition system at the moment k of the target through the detection algorithm p _k =[p _x k ,p _yk ], get the relative angle of the target relative to the camera: ${\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; the\; tan</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}\frac{{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; w</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; \&mu;</span>}}$ ${\mathrm{<span\; class="notranslate">\; \&gamma;</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; the\; tan</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}\frac{{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; \&mu;</span>}}$ Among them, w and h are the width and height of the image in pixels, respectively, f is the focal length of the camera lens, and μ is the pixel size; According to the installation position of the camera, the relative angle of the target relative to the UAV at moment k can be obtained as Θ _k = [σ _k ,γ _k ], where ${\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; =</span>\left\{\begin{array}{cc}{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; 4</span>}}{\mathrm{<span\; class="notranslate">\; 9</span>}}\mathrm{<span\; class="notranslate">\; \&pi;</span>}<span\; class="notranslate">\; ,</span>& {\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; \&Subset;</span>\mathrm{<span\; class="notranslate">\; A</span>}\\ {\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; ,</span>& {\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; \&Subset;</span>\mathrm{<span\; class="notranslate">\; B</span>}\\ {\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 4</span>}}{\mathrm{<span\; class="notranslate">\; 9</span>}}\mathrm{<span\; class="notranslate">\; \&pi;</span>}<span\; class="notranslate">\; ,</span>& {\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; \&Subset;</span>\mathrm{<span\; class="notranslate">\; C</span>}\end{array}\right.$ γ _k = γ′ In the formula Respectively indicate that the detected target comes from camera A, camera B and camera C in the image acquisition system; Determine the angular velocity of the target Ω _k =[ω _xk ,ω _yk ]=Θ _k+1 -Θ _k ,k>0 according to the angle changes at the front and rear moments; Step 3, according to the state information of the flight target obtained in the step 2, the safety envelope module in the visual servo control system performs image plane angle safety envelope generation, and determines whether the UAV has a collision threat through the safety envelope, and aims at the collision threat , calculate the avoidance angle in the image sense; Step 4, input the evasive angle in the step 3 into the visual servo in the visual servo control system, the visual servo outputs the required evasive maneuver pitch and yaw angle, and transmits to the UAV autopilot to complete the evasive maneuver; Step five, use the navigation and positioning system to establish a route return mechanism, and when there is no threat to the drone, make the drone return to the original route. 4. avoidance method according to claim 3, is characterized in that, the generation step of safety envelope and the calculating step of avoidance angle in described step 3 comprise: ①Establish the target safety enveloping circle, with the center of the circle as the target point, select an initial value of the radius, and grow at a fixed speed; ②Create a safe envelope ellipse according to the size of the target. The center of the ellipse is the target point. The length of the minor axis of the ellipse is equal to the radius of the envelope circle. The sum of the extrapolated lengths of , the long axis direction is parallel to the moving direction of the target; ③ The aforementioned generated envelope circle and envelope ellipse form a plane angle safety envelope, which is composed of a semi-ellipse along the direction of target motion and a semi-circle along the opposite direction of target motion with the short axis of the ellipse as the dividing line; ④ Estimate the target threat based on the aforementioned plane angle safety envelope, and establish a threat marking function based on the safety envelope to judge whether there is a collision threat; ⑤ According to the principle of minimum avoidance distance, find the point closest to the initial target in the plane angle safety envelope. 5. avoidance method according to claim 4, is characterized in that, the concrete generation step of safety envelope and the concrete calculation procedure of avoidance angle in described step 3 are as follows: ①Establish the target safety enveloping circle, select the center of the circle as the relative angular position Θ _k ＝[σ _k ,γ _k ] of the target relative to the UAV at moment k, select an initial value r ₀ of the radius, and increase it at a fixed speed, then r _k+1 =r _k +ε,k=1,2,3... r _k = r ₀ , k = 0 Where ε is the fixed growth length of the radius of the safety envelope circle, and r _k+1 represents the radius of the safety envelope circle at time k+1; ② Establish a safety envelope ellipse according to the size of the target, the center of the ellipse is the target point Θ _k = [σ _k , γ _k ], the long axis direction is parallel to the target’s movement direction, that is, the direction of the target angular velocity Ω _k , the long axis and the short axis The length is selected according to the following formula: a=r _k (1+α‖Ω _k ‖), b=r _k , In the formula, a and b represent the major axis and minor axis of the envelope ellipse respectively, r _k is the radius of the envelope circle at time k, α is the scaling factor for converting angular velocity into axis length, which represents the reserved time for avoidance, ‖Ω _k ‖ is the modulus of the target motion angular velocity vector; ③ The aforementioned generated envelope circle and envelope ellipse form a plane angle safety envelope, which is composed of a semi-ellipse along the direction of target motion and a semi-circle along the opposite direction of target motion with the short axis of the ellipse as the dividing line; ④ Estimate the target threat based on the above-mentioned plane angle safety envelope, and establish a threat marking function l _k based on time k. When l _k <0, there is a collision threat, so as to judge whether there is a collision threat, where ${\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; =</span>\left\{\begin{array}{cc}\mathrm{<span\; class="notranslate">\; sgn</span>}<span\; class="notranslate">\; (</span>{<span\; class="notranslate">\; (</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; \&gamma;</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&gamma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>& <span\; class="notranslate">\; (</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; \&Omega;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; 0</span>}\\ \mathrm{<span\; class="notranslate">\; sgn</span>}<span\; class="notranslate">\; (</span>\frac{{<span\; class="notranslate">\; (</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; +</span>\frac{{<span\; class="notranslate">\; (</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; \&gamma;</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&gamma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>& <span\; class="notranslate">\; (</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; \&Omega;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; \&Greater\; Equal;</span>\mathrm{<span\; class="notranslate">\; 0</span>}\end{array}\right.<span\; class="notranslate">\; ,</span>$ in Indicates the angle of the main axis direction of camera B relative to the main axis direction of the drone, where, sgn( ) is a symbolic function; ⑤When l _k <0, according to the principle of the minimum avoidance distance, find the distance in the plane angle safety envelope The nearest point s ^* = [σ ^* ,γ ^* ] as the minimum avoidance maneuver point: ${\mathrm{<span\; class="notranslate">\; the\; s</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; =</span>\left\{\begin{array}{cc}{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; +</span>\frac{<span\; class="notranslate">\; (</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; ,</span>& <span\; class="notranslate">\; (</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; \&Omega;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; 0</span>}\\ {\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; \&PlusMinus;</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; \&perp;</span>{\mathrm{<span\; class="notranslate">\; \&Omega;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; ,</span>& <span\; class="notranslate">\; (</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&Theta;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; \&Omega;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; \&Greater\; Equal;</span>\mathrm{<span\; class="notranslate">\; 0</span>}\end{array}\right.<span\; class="notranslate">\; ,</span>$ in Represents the unit vector perpendicular to the target angular velocity vector at time k. 6. The evasion method according to claim 3, characterized in that the specific calculation steps of the evasive maneuver pitch and yaw angle in the step 4 are as follows: ① Define the interaction matrix as ${\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}}<span\; class="notranslate">\; =</span>\left[\begin{array}{cccccc}\frac{<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; cos\&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; cos\&gamma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}{\mathrm{<span\; class="notranslate">\; d</span>}}& \frac{<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; cos\&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; sin\&gamma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}{\mathrm{<span\; class="notranslate">\; d</span>}}& {\mathrm{<span\; class="notranslate">\; sin\&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}& {\mathrm{<span\; class="notranslate">\; sin\&gamma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}& <span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; cos\&gamma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}& \mathrm{<span\; class="notranslate">\; 0</span>}\\ \frac{{\mathrm{<span\; class="notranslate">\; sin\&gamma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}{\mathrm{<span\; class="notranslate">\; d</span>}{\mathrm{<span\; class="notranslate">\; sin\&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}& \frac{<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; cos\&gamma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}{\mathrm{<span\; class="notranslate">\; d</span>}{\mathrm{<span\; class="notranslate">\; sin\&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}& \mathrm{<span\; class="notranslate">\; 0</span>}& \frac{{\mathrm{<span\; class="notranslate">\; cos\&gamma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; cos\&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}{{\mathrm{<span\; class="notranslate">\; sin\&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}& \frac{{\mathrm{<span\; class="notranslate">\; sin\&gamma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; cos\&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}{{\mathrm{<span\; class="notranslate">\; sin\&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}& <span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}\end{array}\right]<span\; class="notranslate">\; ,</span>$ The speed controller is established in the visual servo controller of the visual servo control system, and the feedback amount is The exponential attenuation of the error is completed, and the output speed control value V is expressed as $\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; \&rsqb;</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;L</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; +</span>}\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; ,</span>$ In the formula, λ is the error exponential decay coefficient, Represents the pseudo-inverse of L _e ; Since the information of the distance d between the target and the machine cannot be obtained in the interaction matrix L _e , the interaction matrix is divided as follows: ${\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; d</span>}}{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}<span\; class="notranslate">\; \&rsqb;</span><span\; class="notranslate">\; ,</span>$ In the formula, L _ω represents the angle control component in L _e , and L _t represents the speed motion control component; Get a distance-independent visual servo controller: $\mathrm{<span\; class="notranslate">\; \&omega;</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;L</span>}}_{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}^{<span\; class="notranslate">\; +</span>}\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; ,</span>$ In the formula, represents the pseudo-inverse of L _ω ; In order to eliminate the steady-state error, the visual servo controller is compensated as follows: $\mathrm{<span\; class="notranslate">\; \&omega;</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;L</span>}}_{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}^{<span\; class="notranslate">\; +</span>}\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}^{<span\; class="notranslate">\; +</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&Omega;</span>}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; s</span>}}^{<span\; class="notranslate">\; *</span>}}{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; the\; s</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}\mathrm{<span\; class="notranslate">\; \&epsiv;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\left[\begin{array}{ccc}{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}& {\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}}& {\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}\end{array}\right]<span\; class="notranslate">\; ,</span>$ In the formula, Ω represents the angular velocity of the target movement, Indicates the direction vector of the error, ω _x , ω _y , ω _z represent the three components of ω respectively; ②The output pitch angle and yaw angle are ${\mathrm{<span\; class="notranslate">\; \&phi;</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&phi;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\frac{\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; g</span>}}$ In the formula, and φ _c (k) represent the pitch and yaw angles of the autopilot output attitude, and φ(k) represent the pitch angle and yaw angle of the current attitude of the UAV, Δt represents the output time interval step of the visual servo controller, ω _y (k) and ω _z (k) represent the visual servo controller at the current moment The output pitch rate control amount and yaw rate control amount, g represents the acceleration of gravity, and V(k) represents the speed of the drone at the current moment. 7. The avoidance method according to claim 3, characterized in that, the specific steps of establishing the route return mechanism in the step 5 are as follows: ① Determine whether the UAV is on the route through the local navigation and positioning system, and return to the route when the aircraft is not on the route; ②Confirm the return route angle at time k ${\mathrm{<span\; class="notranslate">\; \&gamma;</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; cos</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}\frac{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; 12</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; 12</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; 12</span>}}}{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; 12</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; 12</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; \&pi;</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ,</span>$ ${\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; cos</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}\frac{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; 13</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; 13</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; 13</span>}}}{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; 13</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; 13</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 3</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}{\mathrm{<span\; class="notranslate">\; 4</span>}}<span\; class="notranslate">\; ,</span>$ In the formula, Represents the position of the minimum maneuver avoidance point when the route returns; Where P _d is the position of the own aircraft on the original route when there is no evasive maneuver, P _d12 and P _d13 are the two-dimensional positions of the own aircraft on the XY and XZ planes respectively; P _s is the current position of the machine obtained through the local navigation and positioning system, and P _s12 and P _s13 are the two-dimensional positions of the machine on the XY and XZ planes respectively; e _s is the local direction vector obtained by the local navigation and positioning system, and e _s12 and e _s13 are the two-dimensional direction vectors of the local plane on the XY and XZ planes, respectively.