Air-ground cooperative region specific target searching method

Abstract

The invention discloses an air-ground cooperative area specific target searching method which is completed through a control center, and an unmanned vehicle and an unmanned aerial vehicle which are wirelessly connected with the control center. The specific searching method comprises the following steps: (1) loading goods and materials on the unmanned vehicle, and loading the unmanned aerial vehicle on the take-off and landing platform; (2) delimiting a search area formed by four longitude and latitude coordinate points, and sending the four coordinate points to the unmanned vehicle through the control center. The invention provides an air-ground cooperative region specific target searching method, which can effectively enable an unmanned vehicle to be linked with an unmanned aerial vehicle, can effectively utilize the excellent three-dimensional space moving capability and view advantage of the unmanned aerial vehicle, can effectively utilize the load and cruising ability of the unmanned vehicle to improve the searching radius of the unmanned aerial vehicle, and improves the searching efficiency of the unmanned aerial vehicle. The unmanned aerial vehicle and the unmanned vehicle complement each other through a linkage mode, so that a specific target can be searched quickly and accurately.

Description

Technical Field

[0001] The present application relates to the field of unmanned aerial vehicles, and more specifically to an air-ground coordinated regional specific target search method. Background Art

[0002] In the field of robotics, the research on multi-robot systems has always been a hot topic. Many universities have invested in the research of robot swarm systems. Among them, swarm applications such as drone formation flight performances, campus transportation unmanned vehicle swarms, and unmanned boat swarm water monitoring have entered the public's field of vision, but the demand for multi-robot swarm systems is far more than that. As technology develops, everyone realizes the power of robot swarm systems, but also recognizes the shortcomings of current swarm systems. At present, most of the research on swarm systems is still stuck in swarm systems composed of a single robot, but many application scenarios cannot be met by a single type of robot swarm system.

[0003] For example, in application scenarios such as forest fire fighting and rescue, inspection and search and rescue of drowning personnel, and target search and capture, a swarm system composed of a single drone, unmanned vehicle, or unmanned boat cannot complete the task well. A swarm system that combines the advantages of various types of robots can achieve the effect of 1+1>2. Combining the excellent three-dimensional spatial mobility and field of view of drones with the excellent load-bearing capacity of unmanned vehicles can fully meet the needs of application scenarios such as forest fire fighting and rescue, and target search and capture; combining the characteristics of drones and unmanned boats can also meet the needs of application scenarios such as inspection and search and rescue of drowning personnel and marine garbage collection.

[0004] Research on heterogeneous cross-domain robot cluster systems such as air-ground collaboration, sea-air collaboration, and sea-land-air collaboration will be the focus of future cluster development research. Summary of the invention

[0005] In response to the shortcomings of the existing technology, the present application provides an air-ground collaborative regional specific target search method, which can effectively enable unmanned vehicles and drones to work together, effectively utilizing the drone's excellent three-dimensional spatial mobility and field of view advantages, and effectively utilizing the unmanned vehicle's load capacity and endurance to increase the drone's search radius. Through the linkage method, the drone and unmanned vehicle can complement each other, thereby quickly and accurately completing the search for specific targets.

[0006] A method for searching for specific targets in an air-ground collaborative area is implemented by a control center, and an unmanned vehicle and an unmanned aerial vehicle wirelessly connected to the control center; the unmanned vehicle and the unmanned aerial vehicle are respectively equipped with communication modules that enable them to interact with each other; the unmanned vehicle is also equipped with an on-board computing unit, a positioning module and a sensor unit that enable it to perceive its own posture, acquire environmental data and perform autonomous analysis and calculation; the unmanned vehicle is also equipped with a take-off and landing platform for the unmanned aerial vehicle to take off and land; the unmanned aerial vehicle is also equipped with an on-board computing unit, a positioning module and a sensor unit that enable it to perceive its own posture, acquire environmental data and perform autonomous analysis and calculation;

[0007] The specific search method is:

[0008] (1) Load materials onto the unmanned vehicle and load the drone onto the take-off and landing platform;

[0009] (2) Delineate a search area formed by four longitude and latitude coordinate points, and send the four coordinate points to the unmanned vehicle through the control center;

[0010] (3) The onboard computing unit of the unmanned vehicle generates navigation based on the coordinate points and the preset map and drives to the edge of the search area according to the navigation;

[0011] (4) The unmanned vehicle stops after driving to the perimeter of the search area, and the vehicle-mounted computing unit sends four longitude and latitude coordinates and a take-off command to the airborne computing unit;

[0012] (5) The UAV takes off, and the onboard computing unit creates a rectangular map with the four longitude and latitude coordinate points as vertices based on the four longitude and latitude coordinate points;

[0013] (6) The drone moves in the created rectangular map and takes images while moving;

[0014] (7) The onboard computing unit detects the image in real time. If a specific target is detected, the process proceeds to step (8). If the specific target is still not detected after the rectangular map search is completed, the onboard computing unit controls the UAV to return to the lifting platform.

[0015] (8) The onboard computing unit obtains the location information of the specific target and sends the location information of the specific target to the onboard computing unit;

[0016] (9) The on-board computing unit generates navigation according to the location information of the specific target, and travels to the location of the specific target according to the navigation.

[0017] In step (6), the UAV's travel path is the local optimal path in the map obtained by calculating the MPC algorithm and the particle swarm algorithm. The optimal path refers to the undetected area that the UAV has traveled the most and does not exceed the map boundary.

[0018] The calculation process of the optimal path in step (6) is achieved by establishing a motion model prediction, specifically:

[0019] (61) The UAV is regarded as a particle and a particle model is established to represent it. The position prediction equation of the particle is:

[0020]

[0021] The angle prediction equation for this particle is:

[0022]

[0023] Among them, x i (k+1) and y i (k+1) is the prediction of the x-axis position and y-axis position in the map coordinate system at time k+1 at time k, i (k) and y i (k) is the positioning data in the map system, v0 is the control speed, Δt is the inverse of the control frequency, that is, the control interval, is the velocity direction in the map coordinate system at time k, is the prediction of the angle at time k, u is the control amount, that is, the expected increment of speed, e x is the X-axis perturbation, e y is the Y-axis disturbance;

[0024] (62) The control data of the UAV and the actual flight data output are collected as data sets. The actual position of the next moment in the map coordinate system under the action of u is fitted using a neural network to complete the modeling of the motion model. The fitted linear model is:

[0025] val(x,y)=p00+p10*x+p20*x 2 +p11*x*y+p02*y 2 +p30*

[0026] x 3 +p21*x 2 *y+p12*x*y 2 +p03*y 3 +p40*x 4 +p31*x 3 *y+p22*

[0027] x 2 *y 2 +p13*x*y 3 +p04*y 4 +p50*x 5 +p41*x 4 *y+p32*x3 *y 2 +

[0028] p23*x 2 *y 3 +p14*x*y 4 +p05*y 5 ,

[0029] Among them, p is the fitting coefficient, x is the current position, and y is the component of the current control amount in the map coordinate system;

[0030] The specific fit coefficients with a 95% confidence limit are:

[0031] p00=-0.0175(-0.06599,0.03098),

[0032] p10=1.016(0.9595,1.072),

[0033] p01=0.9137(0.8279,0.9995),

[0034] p20=0.02251(-0.02736,0.07239),

[0035] p11=0.1017(0.02979,0.1735),

[0036] p02=0.05081(-0.01673,0.1184),

[0037] p30=-0.01165(-0.03908,0.01577),

[0038] p21=-0.05525(-0.114,0.003526),

[0039] p12=0.007401(-0.06099,0.07579),

[0040] p03=-0.3432(-0.3887,-0.2978),

[0041] p40=-0.004387(-0.01595,0.007178),

[0042] p31=-0.02174(-0.04617,0.002696),

[0043] p22=0.004047(-0.01587,0.02396),

[0044] p13=-0.01926(-0.03133,-0.007195),

[0045] p04=-0.01899(-0.02833,-0.009649),

[0046] p50=0.001632(-0.002588,0.005852),

[0047] p41=0.009203(-0.001797,0.0202),

[0048] p32=0.0008695(-0.01035,0.01209),

[0049] p23=0.01341(0.002299,0.02452),

[0050] p14=-0.007415(-0.01781,0.002976),

[0051] p05=0.02513(0.02006,0.0302);

[0052] Substitute the fitting coefficients into the linear model. In order to improve the calculation efficiency, the linear model finally obtained after retaining the first four items is:

[0053] val(x,y)=-0.0175+1.016*x(k)+0.9137*cos(u)+0.1*x(k)*cos(u);

[0054] (63) Substituting the linear model into the formula of step (61), we finally obtain:

[0055] Position prediction equation:

[0056]

[0057] Angle prediction equation:

[0058]

[0059] In step (8), the method for obtaining the location information of the specific target is:

[0060] (81) After the onboard computing unit detects the specific target, it constructs a camera coordinate system and marks the position of the specific target in the camera coordinate system;

[0061] (82) The onboard computer constructs a UAV coordinate system, and obtains the position of the specific target in the UAV coordinate system through TF relationship transformation according to the position of the specific target in the camera coordinate system;

[0062] (83) The onboard computer obtains GPS information of the drone and the unmanned vehicle;

[0063] (84) The onboard computer converts the GPS information of the UAV and the unmanned vehicle into coordinates in the ECEF coordinate system, and obtains the position distance between the UAV and the unmanned vehicle in the ECEF coordinate system;

[0064] (85) Calculate the deviation based on the position distance between the UAV and the UAV in the ECEF coordinate system, and convert the deviation into the deviation in the northeast sky coordinate system;

[0065] (86) creating a local map and a local map coordinate system according to the local coordinates of the unmanned vehicle, and converting the northeast sky coordinate system into the local map coordinate system;

[0066] The local coordinates are the coordinate points of the drone's location when it takes off.

[0067] (87) respectively obtaining the coordinate points of the unmanned vehicle and the specific target in the local map coordinate system;

[0068] (88) The onboard computing unit sends the local map coordinate system, the coordinate points of the unmanned vehicle on the local map coordinate system, and the coordinate points of the specific target on the local map coordinate system to the onboard computing unit.

[0069] In step (9), the unmanned vehicle collects images in real time during its movement to determine whether the path is passable. If not, a new navigation is generated in real time until the unmanned vehicle reaches the location of the specific target.

[0070] In step (9), if the unmanned vehicle cannot travel to the specific target location according to the generated navigation, the unmanned vehicle sends feedback information to the control center.

[0071] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0072] The present invention can effectively link the unmanned vehicle with the drone, which can effectively utilize the excellent three-dimensional spatial mobility and field of view of the drone, and can effectively utilize the load capacity and endurance of the unmanned vehicle to increase the search radius of the drone. Through linkage, the drone and the unmanned vehicle can complement each other, thereby quickly and accurately completing the search for specific targets.

[0073] Some additional features of the present application may be explained in the following description. Some additional features of the present application will be apparent to those skilled in the art by inspection of the following description and corresponding drawings or understanding of the production or operation of the embodiments. The features disclosed in the present application may be realized and achieved by practice or use of various methods, means and combinations of the specific embodiments described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The exemplary embodiments of the present application and their descriptions are used to explain the present application and do not constitute a limitation of the present application. In each figure, the same reference numeral represents the same component.

[0075] Figure 1 It is a flowchart of the implementation steps of the present invention.

[0076] Figure 2 4 is a neural network diagram of an embodiment of the present invention.

[0077] Figure 3 It is the residual graph of val(x, y) in the embodiment of the present invention. DETAILED DESCRIPTION

[0078] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0079] It should be noted that if the description and claims of the present application and the above-mentioned drawings relate to the terms "first", "second", etc., they are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, if the terms "including" and "having" and any of their variations are involved, it is intended to cover non-exclusive inclusions, for example, the process, method, system, product or equipment that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0080] In this application, if the terms "upper", "lower", "left", "right", "front", "back", "top", "bottom", "inside", "outside", "middle", "vertical", "horizontal", "lateral", "longitudinal", etc. are involved, the orientation or position relationship indicated is based on the orientation or position relationship shown in the drawings. These terms are mainly used to better describe the present application and its embodiments, and are not used to limit the indicated devices, elements or components to have a specific orientation, or to be constructed and operated in a specific orientation.

[0081] In addition, some of the above terms may be used to express other meanings in addition to indicating orientation or positional relationship. For example, the term "on" may also be used to express a certain dependency or connection relationship in some cases. For those of ordinary skill in the art, the specific meanings of these terms in this application can be understood according to specific circumstances.

[0082] In addition, in this application, the terms "installed", "set", "provided with", "connected", "connected", "socketed" and the like should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral structure; it can be a mechanical connection, or an electrical connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be an internal connection between two devices, elements or components. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to the specific circumstances.

[0083] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0084] Example 1

[0085] A method for searching for specific targets in an air-ground collaborative area is implemented by a control center, and an unmanned vehicle and an unmanned aerial vehicle wirelessly connected to the control center; the unmanned vehicle and the unmanned aerial vehicle are respectively equipped with communication modules that enable them to interact with each other; the unmanned vehicle is also equipped with an on-board computing unit, a positioning module and a sensor unit that enable it to perceive its own posture, acquire environmental data and perform autonomous analysis and calculation; the unmanned vehicle is also equipped with a take-off and landing platform for the unmanned aerial vehicle to take off and land; the unmanned aerial vehicle is also equipped with an on-board computing unit, a positioning module and a sensor unit that enable it to perceive its own posture, acquire environmental data and perform autonomous analysis and calculation;

[0086] like Figure 1 As shown, the specific search method is:

[0087] (1) Load materials onto the unmanned vehicle and load the drone onto the take-off and landing platform;

[0088] (2) Delineate a search area formed by four longitude and latitude coordinate points, and send the four coordinate points to the unmanned vehicle through the control center;

[0089] (3) The onboard computing unit of the unmanned vehicle generates navigation based on the coordinate points and the preset map and drives to the edge of the search area according to the navigation;

[0090] (4) The unmanned vehicle stops after driving to the perimeter of the search area, and the vehicle-mounted computing unit sends four longitude and latitude coordinates and a take-off command to the airborne computing unit;

[0091] (5) The UAV takes off, and the onboard computing unit creates a rectangular map with the four longitude and latitude coordinate points as vertices based on the four longitude and latitude coordinate points;

[0092] (6) The drone moves in the created rectangular map and takes images while moving;

[0093] The UAV's travel path is the local optimal path in the map calculated by the MPC algorithm and the particle swarm algorithm. The optimal path refers to the undetected area that the UAV has traveled the most and does not exceed the map boundary.

[0094] The calculation process of the optimal path is achieved by establishing a motion model prediction, specifically:

[0095] (61) The UAV is regarded as a particle and a particle model is established to represent it. The position prediction equation of the particle is:

[0096]

[0097] The angle prediction equation for this particle is:

[0098]

[0099] Among them, x i (k+1) and y i (k+1) is the prediction of the x-axis position and y-axis position in the map coordinate system at time k+1 at time k, i (k) and y i (k) is the positioning data in the map system, v0 is the control speed, Δt is the inverse of the control frequency, that is, the control interval, is the velocity direction in the map coordinate system at time k, is the prediction of the angle at time k, u is the control amount, that is, the expected increment of speed, e x is the X-axis perturbation, e y is the Y-axis disturbance;

[0100] (62) The control data and actual flight data of the UAV are collected as a data set. The actual position of the next moment in the map coordinate system under the action of u is fitted using a neural network to complete the modeling of the motion model. The neural network is as follows: Figure 2 A two-layer feedforward network with sigmold hidden neurons and linear output neurons, suitable for regression tasks, is shown. The neural network is trained to map predictor variables to continuous responses.

[0101] The relevant parameters of the neural network are set as follows:

[0102] data:

[0103] Predictor variables: data5-[387*2double];

[0104] Corresponding: data2-[387*1double];

[0105] Among them, data5 is a double array consisting of 387 observations with 2 features, and data2 is a double array consisting of 387 observations with 1 feature.

[0106] algorithm:

[0107] Data partitioning: random;

[0108] Training algorithm: Levenberg-Marquardt method;

[0109] Performance: Mean squared error.

[0110] The fitted linear model is:

[0111] val(x,y)=p00+p10*x+p20*x 2 +p11*x*y+p02*y 2 +p30*

[0112] x 3 +p21*x 2 *y+p12*x*y 2 +p03*y 3 +p40*x 4 +p31*x 3 *y+p22*

[0113] x 2 *y 2 +p13*x*y 3 +p04*y 4 +p50*x 5 +p41*x 4 *y+p32*x 3 *y 2 +

[0114] p23*x 2 *y 3 +p14*x*y 4 +p05*y 5 ,

[0115] Among them, p is the fitting coefficient, x is the current position, and y is the component of the current control amount in the map coordinate system;

[0116] The specific fit coefficients with a 95% confidence limit are:

[0117] p00=-0.0175(-0.06599,0.03098),

[0118] p10=1.016(0.9595,1.072),

[0119] p01=0.9137(0.8279,0.9995),

[0120] p20=0.02251(-0.02736,0.07239),

[0121] p11=0.1017(0.02979,0.1735),

[0122] p02=0.05081(-0.01673,0.1184),

[0123] p30=-0.01165(-0.03908,0.01577),

[0124] p21=-0.05525(-0.114,0.003526),

[0125] p12=0.007401(-0.06099,0.07579),

[0126] p03=-0.3432(-0.3887,-0.2978),

[0127] p40=-0.004387(-0.01595,0.007178),

[0128] p31=-0.02174(-0.04617,0.002696),

[0129] p22=0.004047(-0.01587,0.02396),

[0130] p13=-0.01926(-0.03133,-0.007195),

[0131] p04=-0.01899(-0.02833,-0.009649),

[0132] p50=0.001632(-0.002588,0.005852),

[0133] p41=0.009203(-0.001797,0.0202),

[0134] p32=0.0008695(-0.01035,0.01209),

[0135] p23=0.01341(0.002299,0.02452),

[0136] p14=-0.007415(-0.01781,0.002976),

[0137] p05=0.02513(0.02006,0.0302);

[0138] The residual graph of Val(x, y) is as follows Figure 3 shown.

[0139] Substitute the fitting coefficients into the linear model. In order to improve the calculation efficiency, the linear model finally obtained after retaining the first four items is:

[0140] val(x,y)=-0.0175+1.016*x(k)+0.9137*cos(u)+0.1*x(k)*cos(u);

[0141] (63) Substituting the linear model into the formula of step (61), we finally obtain:

[0142] Position prediction equation:

[0143]

[0144] Angle prediction equation:

[0145]

[0146] (7) The onboard computing unit detects the image in real time. If a specific target is detected, the process proceeds to step (8). If the specific target is still not detected after the rectangular map search is completed, the onboard computing unit controls the UAV to return to the lifting platform.

[0147] (8) The onboard computing unit obtains the location information of the specific target and sends the location information of the specific target to the onboard computing unit;

[0148] The method for obtaining the location information of a specific target is:

[0149] (81) After the onboard computing unit detects the specific target, it constructs a camera coordinate system and marks the position of the specific target in the camera coordinate system;

[0150] (82) The onboard computer constructs a UAV coordinate system, and obtains the position of the specific target in the UAV coordinate system through TF relationship transformation according to the position of the specific target in the camera coordinate system;

[0151] (83) The onboard computer obtains GPS information of the drone and the unmanned vehicle;

[0152] (84) The onboard computer converts the GPS information of the UAV and the unmanned vehicle into coordinates in the ECEF coordinate system, and obtains the position distance between the UAV and the unmanned vehicle in the ECEF coordinate system;

[0153] (85) Calculate the deviation based on the position distance between the UAV and the UAV in the ECEF coordinate system, and convert the deviation into the deviation in the northeast sky coordinate system;

[0154] (86) creating a local map and a local map coordinate system according to the local coordinates of the unmanned vehicle, and converting the northeast sky coordinate system into the local map coordinate system;

[0155] (87) respectively obtaining the coordinate points of the unmanned vehicle and the specific target in the local map coordinate system;

[0156] (88) The onboard computing unit sends the local map coordinate system, the coordinate points of the unmanned vehicle on the local map coordinate system, and the coordinate points of the specific target on the local map coordinate system to the onboard computing unit.

[0157] (9) The on-board computing unit generates navigation according to the location information of the specific target, and travels to the location of the specific target according to the navigation.

[0158] The unmanned vehicle collects images in real time while moving to determine whether the path is passable. If not, a new navigation is generated in real time until the unmanned vehicle reaches the location of the specific target.

[0159] If the unmanned vehicle cannot travel to a specific target location according to the generated navigation, the unmanned vehicle sends feedback information to the control center.

[0160] It should be noted that all features disclosed in this specification, or steps in all methods or processes disclosed, except for mutually exclusive features and / or steps, can be combined in any manner.

[0161] In addition, the above specific embodiments are exemplary, and those skilled in the art can come up with various solutions inspired by the disclosure of the present invention, and these solutions also belong to the disclosure scope of the present invention and fall within the protection scope of the present invention. Those skilled in the art should understand that the present invention description and its drawings are illustrative and do not constitute a limitation of the claims. The protection scope of the present invention is defined by the claims and their equivalents.

Claims

1. A method for searching for area-specific targets in air-ground collaboration, characterized in that: The control is completed through a control center and unmanned vehicles and unmanned aerial vehicles wirelessly connected to the control center; the unmanned vehicles and unmanned aerial vehicles are respectively equipped with communication modules that enable them to interact with each other; the unmanned vehicles are also equipped with a vehicle-mounted computing unit, a positioning module and a sensor unit that enable them to perceive their own posture, acquire environmental data and perform autonomous analysis and calculation, and the unmanned vehicles are also equipped with a take-off and landing platform for the unmanned aerial vehicle to take off and land; the unmanned aerial vehicle is also equipped with an airborne computing unit, a positioning module and a sensor unit that enable them to perceive their own posture, acquire environmental data and perform autonomous analysis and calculation; The specific search method is: (1) Load materials onto the unmanned vehicle and load the drone onto the take-off and landing platform; (2) Delineate a search area formed by four longitude and latitude coordinate points, and send the four coordinate points to the unmanned vehicle through the control center; (3) The onboard computing unit of the unmanned vehicle generates navigation based on the coordinate points and the preset map and drives to the edge of the search area according to the navigation; (4) The unmanned vehicle stops after driving to the perimeter of the search area, and the vehicle-mounted computing unit sends four longitude and latitude coordinates and a take-off command to the airborne computing unit; (5) The UAV takes off, and the onboard computing unit creates a rectangular map with the four longitude and latitude coordinate points as vertices based on the four longitude and latitude coordinate points; (6) The drone moves in the created rectangular map and takes images while moving; (7) The onboard computing unit detects the image in real time. If a specific target is detected, the process proceeds to step (8). If the specific target is still not detected after the rectangular map search is completed, the onboard computing unit controls the UAV to return to the lifting platform. (8) The onboard computing unit obtains the location information of the specific target and sends the location information of the specific target to the onboard computing unit; (9) The on-board computing unit generates navigation according to the location information of the specific target, and travels to the location of the specific target according to the navigation.

2. The method for searching for area-specific targets in air-ground collaboration according to claim 1, characterized in that: In step (6), the UAV's travel path is the local optimal path in the map obtained by calculating the MPC algorithm and the particle swarm algorithm. The optimal path refers to the undetected area that the UAV has traveled the most and does not exceed the map boundary.

3. The method for searching for area-specific targets in air-ground collaboration according to claim 2, characterized in that: The calculation process of the optimal path in step (6) is achieved by establishing a motion model prediction, specifically: (61) The UAV is regarded as a particle and a particle model is established to represent it. The position prediction equation of the particle is: The angle prediction equation for this particle is: Among them, x i (k+1) and y i (k+1) is the prediction of the x-axis position and y-axis position in the map coordinate system at time k+1 at time k, i (k) and y i (k) is the positioning data in the map system, v0 is the control speed, Δt is the inverse of the control frequency, that is, the control interval, is the velocity direction in the map coordinate system at time k, is the prediction of the angle at time k, u is the control amount, that is, the expected increment of speed, e x is the X-axis perturbation, e y is the Y-axis disturbance; (62) The control data of the UAV and the actual flight data output are collected as data sets. The actual position of the next moment in the map coordinate system under the action of u is fitted using a neural network to complete the modeling of the motion model. The fitted linear model is: val(x,y)=p00+p10*x+p20*x 2 +p11*x*y+p02*y 2 +p30* x 3 +p21*x 2 *y+p12*x*y 2 +p03*y 3 +p40*x 4 +p31*x 3 *y+p22* x 2 *y 2 +p13*x*y 3 +p04*y 4 +p50*x 5 +p41*x 4 *y+p32*x 3 *y 2 + p23*x 2 *y 3 +p14*x*y 4 +p05*y 5 , Among them, p is the fitting coefficient, x is the current position, and y is the component of the current control amount in the map coordinate system; The specific fit coefficients with a 95% confidence limit are: p00＝-0.0175(-0.06599,0.03098)， p10＝1.016(0.9595,1.072)， p01＝0.9137(0.8279,0.9995)， p20＝0.02251(-0.02736,0.07239)， p11＝0.1017(0.02979,0.1735)， p02＝0.05081(-0.01673,0.1184)， p30＝-0.01165(-0.03908,0.01577)， p21＝-0.05525(-0.114,0.003526)， p12＝0.007401(-0.06099,0.07579)， p03＝-0.3432(-0.3887，-0.2978)， p40＝-0.004387(-0.01595,0.007178)， p31＝-0.02174(-0.04617，0.002696)， p22＝0.004047(-0.01587,0.02396)， p13＝-0.01926(-0.03133,-0.007195)， p04＝-0.01899(-0.02833，-0.009649)， p50＝0.001632(-0.002588,0.005852)， p41＝0.009203(-0.001797,0.0202)， p32＝0.0008695(-0.01035,0.01209)， p23＝0.01341(0.002299，0.02452)， p14＝-0.007415(-0.01781,0.002976)， p05＝0.02513(0.02006,0.0302)； Substitute the fitting coefficients into the linear model. In order to improve the calculation efficiency, the linear model finally obtained after retaining the first four items is: val(x,y)=-0.0175+1.016*x(k)+0.9137*cos(u)+0.1*x(k)*cos(u); (63) Substituting the linear model into the formula of step (61), we finally obtain: Position prediction equation: Angle prediction equation:

4. The method for searching for area-specific targets in air-ground collaboration according to claim 3 is characterized in that: In step (8), the method for obtaining the location information of the specific target is: (81) After the onboard computing unit detects the specific target, it constructs a camera coordinate system and marks the position of the specific target in the camera coordinate system; (82) The onboard computer constructs a UAV coordinate system, and obtains the position of the specific target in the UAV coordinate system through TF relationship transformation according to the position of the specific target in the camera coordinate system; (83) The onboard computer obtains GPS information of the drone and the unmanned vehicle; (84) The onboard computer converts the GPS information of the UAV and the unmanned vehicle into coordinates in the ECEF coordinate system, and obtains the position distance between the UAV and the unmanned vehicle in the ECEF coordinate system; (85) According to the position distance between the UAV and the UAV in the ECEF coordinate system, the deviations of the UAV and the UAV relative to the origin are calculated respectively, and the deviations are converted into deviations in the northeast sky coordinate system; (86) creating a local map and a local map coordinate system according to the local coordinates of the unmanned vehicle, and converting the northeast sky coordinate system into the local map coordinate system; (87) respectively obtaining the coordinate points of the unmanned vehicle and the specific target in the local map coordinate system; (88) The onboard computing unit sends the local map coordinate system, the coordinate points of the unmanned vehicle on the local map coordinate system, and the coordinate points of the specific target on the local map coordinate system to the onboard computing unit.

5. The method for searching for area-specific targets in air-ground collaboration according to claim 4, characterized in that: In step (9), the unmanned vehicle collects images in real time during its movement to determine whether the path is passable. If not, a new navigation is generated in real time until the unmanned vehicle reaches the location of the specific target.

6. The method for searching for area-specific targets in air-ground collaboration according to claim 5, characterized in that: In step (9), if the unmanned vehicle cannot travel to the specific target location according to the generated navigation, the unmanned vehicle sends feedback information to the control center.

Images