CN117434965A - Unmanned aerial vehicle multi-machine collaborative natural pasture intelligent management method and management system - Google Patents
Unmanned aerial vehicle multi-machine collaborative natural pasture intelligent management method and management system Download PDFInfo
- Publication number
- CN117434965A CN117434965A CN202311494514.2A CN202311494514A CN117434965A CN 117434965 A CN117434965 A CN 117434965A CN 202311494514 A CN202311494514 A CN 202311494514A CN 117434965 A CN117434965 A CN 117434965A
- Authority
- CN
- China
- Prior art keywords
- grazing
- unmanned aerial
- aerial vehicle
- pasture
- herd
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000007726 management method Methods 0.000 title claims abstract description 38
- 238000009304 pastoral farming Methods 0.000 claims abstract description 162
- 244000144980 herd Species 0.000 claims abstract description 119
- 238000000034 method Methods 0.000 claims abstract description 75
- 238000013439 planning Methods 0.000 claims abstract description 34
- 230000008569 process Effects 0.000 claims abstract description 31
- 244000144972 livestock Species 0.000 claims description 27
- 238000004422 calculation algorithm Methods 0.000 claims description 23
- 230000033001 locomotion Effects 0.000 claims description 20
- 238000012544 monitoring process Methods 0.000 claims description 15
- 238000012549 training Methods 0.000 claims description 12
- 238000001514 detection method Methods 0.000 claims description 9
- 238000012216 screening Methods 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 8
- 238000007781 pre-processing Methods 0.000 claims description 8
- 238000005520 cutting process Methods 0.000 claims description 7
- 238000012937 correction Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000004891 communication Methods 0.000 claims description 5
- 238000003860 storage Methods 0.000 claims description 4
- 238000013519 translation Methods 0.000 claims description 4
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 claims description 3
- 239000010931 gold Substances 0.000 claims description 3
- 229910052737 gold Inorganic materials 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000013441 quality evaluation Methods 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 238000009795 derivation Methods 0.000 claims description 2
- 230000005484 gravity Effects 0.000 claims description 2
- 230000007306 turnover Effects 0.000 claims description 2
- 230000006870 function Effects 0.000 description 21
- 230000006399 behavior Effects 0.000 description 7
- 230000000694 effects Effects 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 230000002431 foraging effect Effects 0.000 description 5
- 244000025254 Cannabis sativa Species 0.000 description 4
- 238000012876 topography Methods 0.000 description 4
- 230000004888 barrier function Effects 0.000 description 3
- 230000033228 biological regulation Effects 0.000 description 3
- 230000000875 corresponding effect Effects 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 230000002452 interceptive effect Effects 0.000 description 3
- 238000011835 investigation Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 206010000117 Abnormal behaviour Diseases 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000012173 estrus Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000004634 feeding behavior Effects 0.000 description 1
- 230000008014 freezing Effects 0.000 description 1
- 238000007710 freezing Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000010223 real-time analysis Methods 0.000 description 1
- 238000010257 thawing Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/17—Terrestrial scenes taken from planes or by drones
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/005—Tree description, e.g. octree, quadtree
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/05—Geographic models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/292—Multi-camera tracking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/50—Depth or shape recovery
- G06T7/55—Depth or shape recovery from multiple images
- G06T7/593—Depth or shape recovery from multiple images from stereo images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/62—Extraction of image or video features relating to a temporal dimension, e.g. time-based feature extraction; Pattern tracking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/188—Vegetation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
- G06T2207/10012—Stereo images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30188—Vegetation; Agriculture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30232—Surveillance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30241—Trajectory
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Geometry (AREA)
- Human Computer Interaction (AREA)
- Remote Sensing (AREA)
- Evolutionary Computation (AREA)
- Computer Graphics (AREA)
- Psychiatry (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Social Psychology (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention discloses an intelligent management method and system for a natural pasture with multiple unmanned aerial vehicles cooperated, and belongs to the technical field of unmanned aerial vehicles. The method comprises the following steps: acquiring pasture environment information through an unmanned aerial vehicle group; the unmanned aerial vehicle group automatically plans a global grazing path according to the pasture environment information, the obtained user-defined grazing time and the starting point and the ending point, and automatically drives grazing; in the automatic driving grazing process based on the global grazing path, acquiring and analyzing movable images of the herd, planning a local grazing path, and assisting in automatic driving grazing; and in the automatic driving grazing process based on the global grazing path, collecting and analyzing vegetation coverage image data of the grazing area to assist in automatic driving grazing. According to the invention, autonomous grazing of the unmanned aerial vehicle can be realized without the operation of herder.
Description
Technical Field
The invention relates to the technical field of unmanned aerial vehicles, in particular to an intelligent management method and system for a natural pasture with multiple unmanned aerial vehicles in cooperation.
Background
In recent years, unmanned aerial vehicles are widely used as a convenient low-altitude near-ground remote sensing platform with low cost in various fields such as land resource investigation, wild animal investigation and the like. Compared with the traditional aviation aircraft investigation, the unmanned aerial vehicle is more flexible and quiet, and can effectively avoid interfering with daily activities of the herd while keeping a proper supervision distance with the herd. The unmanned aerial vehicle is applied to herd grazing of natural pastures, and is beneficial to achieving accurate management of the herd, individual identification and the like in real time and high efficiency.
However, how to realize tracking grazing of the herd, and rotate grazing areas according to the coverage condition of the vegetation of the pasture, and when the herd normally advances or forges, the animal swarm moving images are obtained through the unmanned aerial vehicle to realize individual behavior detection of the livestock, so that the accurate grazing management is a problem to be solved.
Disclosure of Invention
In view of the above, the invention provides an unmanned aerial vehicle multi-machine collaborative natural pasture intelligent management method and system, which solve the practical requirements of accurate pasture management and image acquisition monitoring of pasture herds.
In order to achieve the above object, the present invention provides the following technical solutions:
an intelligent management method for a natural pasture with multiple unmanned aerial vehicles in cooperation comprises the following steps:
acquiring pasture environment information through an unmanned aerial vehicle group;
the unmanned aerial vehicle group automatically plans a global grazing path according to the pasture environment information, the obtained user-defined grazing time and the starting point and the ending point, and automatically drives grazing;
in the automatic driving grazing process based on the global grazing path, acquiring and analyzing movable images of the herd, planning a local grazing path, and assisting in automatic driving grazing;
and in the automatic driving grazing process based on the global grazing path, collecting and analyzing vegetation coverage image data of the grazing area to assist in automatic driving grazing.
Optionally, the obtaining pasture environment information by the unmanned aerial vehicle group includes:
acquiring pasture environment image data through a binocular depth camera carried by the unmanned aerial vehicle;
preprocessing the pasture environment image data, and carrying out distortion correction and stereo correction based on a stereo matching method to obtain three-dimensional dense point cloud data;
constructing an octree map based on the three-dimensional dense point cloud data, dynamically adjusting the precision and the resolution, screening out effective three-dimensional dense point cloud data, and constructing a pasture area environment map;
dividing the pasture area environment map into a plurality of areas capable of being alternately grazed, marking obstacles affecting normal flight of the unmanned aerial vehicle group in each area, and storing vegetation coverage conditions of each area to a cloud server to obtain pasture environment information.
Optionally, the storing the vegetation coverage of each area in the cloud server includes: the method comprises the steps of obtaining pasture vegetation image data through an unmanned aerial vehicle group, carrying out quality evaluation screening, smooth noise reduction and image cutting on the pasture vegetation image data, calculating regional vegetation coverage according to the processed pasture vegetation image data, and storing the regional vegetation coverage in a cloud server.
Optionally, the unmanned aerial vehicle group automatically plans a global grazing path according to the pasture environment information, the obtained user-defined grazing time, the starting point and the ending point, and performs automatic driving grazing, including:
acquiring a user-defined grazing starting point and end point region and grazing time;
calling a global path planner based on an A-algorithm to plan a global grazing path according to pasture environment information and the starting point and end point region, wherein the method comprises the following steps:
selecting Manhattan distance between two points of the path as the estimation of the distance between the two points;
the method comprises the steps of acquiring elevation data of terrain through a sensor carried by an unmanned aerial vehicle, establishing a digital elevation model, and adjusting a cost function of a path according to the digital elevation model:
the actual cost function from the starting node to node n is formulated as follows:
G(n)＝G(parent)+Cost(n，parent)；
where the parent node parent is the previous point to node n, G (parent) is the actual Cost from the start point to the parent node, and Cost (n, parent) is the extra Cost from the parent node to node n;
the heuristic estimation cost function H considers the path cost from the node n to the target node gold, and the formula is as follows:
H(n)＝k 1 ×Distance(n，parent)+k 2 x|altitude (n) Altitude (goal) |;
the path cost function is:
F(n)＝G(n)+H(n)
where G (n) and H (n) are the actual cost and the heuristically estimated cost function, the computed path cost function will be used to select the next node to be expanded and generate the global grazing path.
Optionally, the calculation formula of the extra Cost (n, parent) from the parent node to the node n is as follows:
Cost(n，parent)＝k 1 ×Distance(n，parent)+k 2
×|Altitude(n)-Altitude(parent)|+k 3
×Speed_Adjusted_Cost(n)+k 4 ×Obstacle_Cost(n)；
wherein Distance (n, parent) is the Manhattan Distance between two points; altitude (n) is the elevation of node n; speed_adjusted_cost (n) is a Cost related to unmanned aerial vehicle flight Speed, including unmanned aerial vehicle flight Speed and herd movement Speed; obstacle_cost (n) is the Cost associated with the obstacles around node n, and is positively related to the density or distance of the obstacles around the path, k 1 、k 2 、k 3 、k 4 Is the weight.
Optionally, in the automatic driving grazing process based on the global grazing path, the method includes collecting and analyzing the movable images of the herd, planning a local grazing path, and assisting the automatic driving grazing process including: in the automatic driving grazing process based on the global grazing path, a movable image of the herd is acquired and analyzed to obtain the position and the motion state of the herd in the actual grazing process, a local path planner based on an improved artificial potential field method is called to plan out a local path of the unmanned aerial vehicle grazing according to the position and the motion state of the herd in the actual grazing process, and the herd is driven to a target grazing area through a stereo player carried by the unmanned aerial vehicle, so that auxiliary automatic driving grazing is realized.
Optionally, the collecting and analyzing the movable images of the herd to obtain the position and the motion state of the herd in the actual grazing process includes:
collecting movable image data of the livestock group through the unmanned aerial vehicle group;
preprocessing the acquired movable image data of the livestock group, including data screening and removing, and performing image rotation, translation, turnover, scaling and cutting processing to generate a data set;
model training, verification, prediction and derivation are carried out on the YOLOv8 target detection model based on the data set, individual identification of the herd and motion trail monitoring and prediction are achieved, and the position and motion state of the herd in the actual grazing process are obtained.
Optionally, the calling the local path planner based on the improved artificial potential field method plans out a local path of the unmanned aerial vehicle grazing, including: the rejection field is expressed using a gaussian function for pointing away from the herd, as follows:
wherein F is repulsion Is the vector of the repulsive force field; n is the number of herd individuals; k is a coefficient for adjusting the intensity of the repulsive force; alpha is a parameter for adjusting the attenuation of the repulsive force; d, d i Is the distance from the unmanned aerial vehicle to the ith herd individual; (x) i ，y i ) Is the location of the ith herd individual; (x, y) is the current position of the drone.
Optionally, the calling the local path planner based on the improved artificial potential field method plans out a local path of the unmanned aerial vehicle grazing, including: the gravity field is expressed by using a Gaussian function and is used for pointing to the direction of the individual herd near the unmanned aerial vehicle (when the individual herd moves, the unmanned aerial vehicle needs to follow monitoring in time), and the formula is as follows:
wherein F is attraction Is the vector of the gravitational field; n is the number of herd individuals; w is a coefficient for adjusting the strength of attraction; beta is a parameter for adjusting the attenuation of the gravitational strength; d, d i Is the distance from the unmanned aerial vehicle to the ith herd individual; (x) i ，y i ) Is the location of the ith herd individual; (x, y) is the current position of the drone.
An unmanned aerial vehicle multi-machine collaborative natural pasture intelligent management system, comprising: the intelligent management system comprises an unmanned aerial vehicle group, a readable storage medium for storing a computer program, a processor for executing the computer program and a cloud server, wherein the unmanned aerial vehicle group consists of a four-rotor unmanned aerial vehicle provided with a sensor, a binocular depth camera, a stereo player and a communication device, and the computer program is executed by the processor to realize the steps of the intelligent management method of the natural pasture coordinated with multiple unmanned aerial vehicles.
According to the technical scheme, compared with the prior art, the invention discloses the intelligent management method and the management system for the natural pasture with the cooperation of multiple unmanned aerial vehicles, and the unmanned aerial vehicle can realize autonomous pasture without operation of herding people. The method has the specific beneficial effects that:
(1) Through unmanned aerial vehicle multimachine cooperation grazing and image acquisition, need not the herd scene and follows, demand such as accessible input custom grazing time, grazing starting point terminal point, grass animal balance threshold value, herd image acquisition, remote control unmanned aerial vehicle system carries out the real-time accurate grazing management of herd according to actual grazing condition, can effectively use manpower and materials to reduce the risk that livestock was lost.
(2) The method comprises the steps of collecting vegetation image data of a current grazing area while grazing, quantitatively evaluating the coverage condition of grassland vegetation, calculating the livestock bearing capacity of a pasture, quickly making a decision of grazing area rotation by combining an environment map, realizing dynamic balance between grassland vegetation production and livestock grazing, efficiently and reasonably utilizing grassland resources, and promoting accurate management of livestock groups and sustainable development of the pasture.
(3) The method has automation and non-invasiveness, can realize real-time grazing and data acquisition and analysis of animal group moving images on the premise of not interfering with normal animal group moving, and further realize real-time monitoring of animal group individuals and the animal group moving images in a normal grazing state, so that abnormal behavior, diseases, oestrus and other conditions of the animal individuals are timely found and uploaded to a system, and the animal individuals can be conveniently known by herds in time; meanwhile, through data acquisition and analysis of the movement track, feeding behavior and the like of livestock individuals, scientific basis can be provided for follow-up accurate grazing management, livestock health is guaranteed, animal welfare is promoted, and production benefits of pastures are improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of the method of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention aims to provide an intelligent management method and system for a natural pasture with multiple unmanned aerial vehicles in cooperation, wherein the method comprises the following steps: acquiring pasture environment information through an unmanned aerial vehicle group; the unmanned aerial vehicle group automatically plans a global grazing path according to the pasture environment information, the obtained user-defined grazing time and the starting point and the ending point, and automatically drives grazing; in the automatic driving grazing process based on the global grazing path, acquiring and analyzing movable images of the herd, planning a local grazing path, and assisting in automatic driving grazing; and in the automatic driving grazing process based on the global grazing path, collecting and analyzing vegetation coverage image data of the grazing area to assist in automatic driving grazing. According to the invention, autonomous grazing of the unmanned aerial vehicle can be realized without the operation of herder.
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
Referring to fig. 1, the embodiment of the invention discloses an unmanned aerial vehicle multi-machine collaborative natural pasture intelligent management method, which comprises the following steps:
acquiring pasture environment information through an unmanned aerial vehicle group;
the unmanned aerial vehicle group automatically plans a global grazing path according to the pasture environment information, the obtained user-defined grazing time and the starting point and the ending point, and automatically drives grazing;
in the automatic driving grazing process based on the global grazing path, acquiring and analyzing movable images of the herd, planning a local grazing path, and assisting in automatic driving grazing;
and in the automatic driving grazing process based on the global grazing path, collecting and analyzing vegetation coverage image data of the grazing area to assist in automatic driving grazing.
In a specific embodiment, before grazing, an unmanned in the system can carry out inspection on surrounding pasture areas for grazing in advance to acquire an environment map. First, pasture environment image data is acquired by a binocular depth camera.
In a specific embodiment, preprocessing is performed on the acquired environment image data, distortion correction and stereo correction are performed on the basis of a stereo matching algorithm, three-dimensional dense point cloud data are obtained, and an environment map is constructed.
In a specific embodiment, considering that the grid map is easy to cause data redundancy, the complexity of map data storage and calculation is increased, the octree map is selected to be constructed based on the obtained three-dimensional dense point cloud data, the precision and resolution can be dynamically adjusted according to the requirement, only meaningful nodes are stored, the data redundancy is reduced, and the complexity of storage and calculation is reduced.
Specifically, the constructed environment map is divided into a plurality of areas which can be grazed in a co-rotation mode according to the number of the herds to be grazed, barriers which possibly affect normal flight of the unmanned aerial vehicle are marked, vegetation coverage conditions of each area are recorded (updated regularly), and the environment map is visible to users. The unmanned aerial vehicle group is according to the grazing time of user definition, the start point terminal point drives the in-process of grazing of herd to involve unmanned aerial vehicle's environment and build the picture, keep away barrier and route planning algorithm. The unmanned aerial vehicle environment map building algorithm is that environment data are collected by using a depth camera sensor carried by the unmanned aerial vehicle, and continuous image frames are subjected to processing such as feature point matching and visual odometer, so that three-dimensional reconstruction of the environment is realized, and an environment map is generated. Because of having many unmanned aerial vehicles, the communication and the data sharing collaborative work between the usable many unmanned aerial vehicles accomplish the environment of the area of grazing jointly and build the map task. The obstacle avoidance algorithm is used for detecting and avoiding obstacles according to sensor data and environmental information, ensuring the safety distance between unmanned aerial vehicles and the safety distance between the unmanned aerial vehicles and the livestock, and avoiding collision between the unmanned aerial vehicles and interference on normal activities of the livestock. The path planning algorithm is used for planning paths of unmanned aerial vehicles in the processes of driving and tracking the herds, image acquisition and the like, and the positions and the motion states of the herds, the relative positions among the unmanned aerial vehicles and obstacle avoidance requirements are required to be considered, so that a proper path is generated, and effective monitoring and image acquisition of the herds are realized.
In a specific embodiment, during grazing, a user can customize a grazing starting point and end point area and grazing time according to an environment map uploaded by an unmanned aerial vehicle system, the unmanned aerial vehicle system calls a global path planner based on an A-algorithm to plan a global path according to the user-defined starting point and end point area, then calls a local path planner based on an improved artificial potential field method to plan a local path of the unmanned aerial vehicle grazing according to the position and the motion state of the livestock in the actual grazing process, and drives the livestock to a target grazing area through a stereo player, and if the livestock are too dispersed, the unmanned aerial vehicle can gather the livestock.
In one embodiment, herd grazing will typically select open areas such as grasslands, rangelands, etc. that are rich in grassland coverage, preferably without significant obstructions or restrictions, with relatively flat terrain, without excessive terrain relief or steep slopes, which may facilitate free movement of the herd. Therefore, the unmanned aerial vehicle system needs to take factors such as large obstacles, topography fluctuation and the like into consideration when constructing an environment map and planning a path, and obtains altitude data of topography through a sensor, such as a Digital Elevation Model (DEM), and the altitude data are integrated into unmanned aerial vehicle path planning.
Specifically, in the algorithm of the global path planning, considering that the path length, the topography fluctuation and the unmanned aerial vehicle flight speed regulation are required to be considered simultaneously in the actual grazing process, the invention improves the path cost function of the algorithm of the global path planning: firstly, considering the length of a planned path, and selecting the Manhattan distance between two points as the estimation of the distance between the two points for saving calculation time; secondly, the cost function of the path is further adjusted according to the collected elevation information, so that the selection of a global grazing path with obvious obstacles or too steep and too high terrain is avoided; finally, considering unmanned aerial vehicle flight speed regulation, flying fast under the condition that no obstacle exists nearby and the herd travels faster, and decelerating to fly under the condition that the obstacle is encountered or the herd moves slower.
The actual cost function from the starting node to node n is formulated as follows:
G(n)＝G(parent)+Cost(n，parent)
where parent node parent is the previous point to node n and G (parent) is the actual cost from the start point to the parent node. Cost (n, parent) is an additional Cost from parent node to node n, formulated as follows:
Cost(n，parent)＝k 1 ×Distance(n，parent)+k 2
×|Altitude(n)-Altitude(parent)|+k 3
×Speed_Adjusted_Cost(n)+k 4 ×Obstacle_Cost(n)
wherein Distance (n, parent) is the Manhattan Distance between two points; altitude (n) is the elevation of node n; speed_adjusted-Cost (n) is a Cost related to unmanned aerial vehicle flight Speed, including unmanned aerial vehicle flight Speed and herd movement Speed; the obstacle_cost (n) is the Cost associated with the obstacles around node n, and is positively correlated with the density or distance of the obstacles around the path. k (k) 1 、k 2 、k 3 、k 4 The weight of the corresponding term can be adjusted according to actual task demands to balance the priority of global path planning in the conditions of path length, topography fluctuation, flying speed regulation and control and surrounding obstacles, and the larger the k value is, the more remarkable the influence of the factor on the path cost is.
The heuristic estimation cost function H considers the path cost from the node n to the target node gold, and the formula is as follows:
H(n)＝k 1 ×Distance(n，parent)+k 2 ×|Altitude(n)-Altitude(goal)|
the path cost function is
F(n)＝G(n)+H(n)
Where G (n) and H (n) are the actual cost and heuristic estimated cost functions defined above. The a algorithm will use this path cost function to select the next node to expand and generate a path.
In a specific embodiment, the local path planning artificial potential field method drives the unmanned aerial vehicle to avoid obstacles and approach targets by introducing potential fields in the environment, however, the conventional artificial potential field method may have some problems in actual grazing scenes, such as target unreachable, local optimum sinking, vibration, and the like. In order to solve the above problems and further improve the effect of unmanned aerial vehicle local path planning, the local path planning artificial potential field method in this embodiment is improved on the basis of the traditional artificial potential field method, and specifically comprises the following improvements: firstly, combining an artificial potential field method with a global path planning method, providing overall path guidance through global path planning, and then using the artificial potential field method to perform local obstacle avoidance so as to overcome the limitation of the artificial potential field method; secondly, dynamically adjusting the shape and strength of a potential field according to the position and motion state of the livestock group, and designing a proper potential field strategy according to the behavior characteristics of the livestock group, such as introducing attractive force and repulsive force to simulate the aggregation and dispersion behaviors of the livestock group so as to better adapt to the actual grazing scene; thirdly, by introducing random disturbance, path re-planning and the like, the unmanned aerial vehicle is prevented from sinking into local optimum, and the robustness of path planning is improved; fourthly, selecting proper sensors, such as a laser radar, a binocular depth camera and the like, sensing the changes of the herd and the environment in real time, and dynamically updating the path planning; and fifthly, taking dynamics and constraint conditions of the unmanned aerial vehicle, such as maximum speed, maximum acceleration and the like, into consideration so as to ensure feasibility and safety of path planning.
Specifically, the artificial potential field method is a path planning method based on a virtual potential field, and the moving object is guided by introducing attractive force and repulsive force fields in the environment so as to achieve the goal of planning a path. In the process of local path planning of unmanned aerial vehicle grazing, corresponding potential field strategies are designed according to the characteristics of animal group behaviors.
Behavioral characteristics of herds typically include: firstly, herds aggregate, which tend to approach each other and form a population; secondly, the herd avoids the obstacle, and the herd can actively avoid the obstacle when advancing. Based on the behavior characteristics, in order to avoid the influence on normal activities of the herd due to the fact that the unmanned aerial vehicle is too close to the herd in flight, an exclusive force field strategy is designed. The magnitude and direction of the repulsive force field should be related to the distance and relative speed between the drone and the herd, and the repulsive force field is expressed using a gaussian function, as follows:
wherein F is repulsion Is the vector of the repulsive force field; n is the number of herd individuals; k is a coefficient for adjusting the intensity of the repulsive force; alpha is a parameter for adjusting the attenuation of the repulsive force; d, d i Is the distance from the unmanned aerial vehicle to the ith herd individual; (x) i ，y i ) Is the location of the ith herd individual; (x, y) is the current position of the drone. This repulsive force field will generate a vector pointing away from the herd and increasing as the distance of the drone from the herd decreases. The specific repulsive force parameters can be adjusted according to grazing experience and task requirements so as to better realize grazing tasks. The unmanned aerial vehicle can be influenced by the repulsive force field in flight so as to ensure that the unmanned aerial vehicle keeps a sufficient distance with the herd and avoid influencing normal activities of the herd.
In addition, when the herd moves, the unmanned aerial vehicle needs to follow monitoring in time, so that a gravitational field strategy is designed to guide the unmanned aerial vehicle to advance along the moving direction of the herd. The magnitude and direction of the gravitational field are also affected by the distance and relative speed between the unmanned aerial vehicle and the herd individuals, and the gravitational field is expressed by a Gaussian function, and the formula is as follows:
wherein F is attraction Is the vector of the gravitational field; n is the number of herd individuals; w is a coefficient for adjusting the strength of attraction; beta is a parameter for adjusting the attenuation of the gravitational strength; d, d i Is the distance from the unmanned aerial vehicle to the ith herd individual; (x) i ,y i ) Is the location of the ith herd individual; (x, y) is the current position of the drone. The gravitational field will generate a vector directed to the individual herd closest to the unmanned aerial vehicle, the strength of the gravitational field will increase as the unmanned aerial vehicle decreases from the herd so that the unmanned aerial vehicle advances in the direction of herd movement. Specific gravitation parameters can be adjusted according to grazing experience and task requirements so as to better realize grazing tasks.
The repulsive force field and the gravitational field are comprehensively considered, potential field parameters are adjusted through a test flight experiment so as to adapt to different animal group behavior characteristics and grazing requirements, and path planning of the unmanned aerial vehicle grazing system in safe driving and guiding of animal groups is achieved.
Specifically, after the improvement of the path planning algorithm is realized, the specific steps for obtaining the grazing route through the path planning algorithm are as follows: firstly, taking a user-defined starting point and end point position coordinate, an environment map obtained by inspection of an unmanned aerial vehicle before grazing and barrier information as inputs of a path planning algorithm; searching a global optimal path from a starting point to an end point based on an improved A-path planning algorithm and a pasture area environment map; thirdly, in the actual flight process of the unmanned aerial vehicle, updating an environment map in real time, and planning a local path under the current condition based on an improved artificial potential field method; fourth, unmanned aerial vehicle execution path.
In a specific embodiment, when the herd is in a stable advancing or foraging state, the system schedules 3 unmanned aerial vehicles to monitor the herd state around, ensures that all the herds are in foraging in a monitoring area, schedules 1-2 unmanned aerial vehicles to fly above the herd, acquires the active image data of the herd and the vegetation coverage condition image data through hovering and overlooking angles, and uploads the active image data and the vegetation coverage condition image data to a server.
In one embodiment, if an individual is out of the monitoring range, 1 unmanned aerial vehicle is dispatched from the monitoring unmanned aerial vehicle group to track the individual, and the unmanned aerial vehicle needs to quickly track the individual out of the monitoring range and drive the individual back to the monitoring range of the livestock range through a stereo player.
In a specific embodiment, the unmanned aerial vehicle system can also perform the image data acquisition work of the movable images of the herd and the vegetation coverage condition of the grazing area while carrying out the collaborative grazing, and the automatic driving grazing is assisted, so that the collaborative work between unmanned aerial vehicles is realized through a multi-machine collaborative algorithm.
Firstly, a plurality of unmanned aerial vehicles in the system are combined into a cooperative cluster so as to realize an unmanned aerial vehicle cluster forming algorithm based on a virtual structure algorithm for intensively driving the herd to a specified grazing place, and the flight track of the unmanned aerial vehicles can be adjusted according to the relative positions and speeds among the unmanned aerial vehicles so as to set geometric shapes or dynamic structures to cooperatively drive and graze the herd.
And secondly, carrying out information exchange and data sharing among unmanned aerial vehicles based on communication and cooperative algorithm of the unmanned aerial vehicle system so as to cooperatively complete grazing and data acquisition tasks.
Then, when the herd is in a stable advancing or foraging state, the system dynamically distributes proper positions and tasks for each unmanned aerial vehicle based on a task distribution algorithm, and cooperatively completes the monitoring grazing, the image data acquisition of the herd state and the environment image data acquisition task of the grazing area so as to realize comprehensive supervision of the herd and real-time monitoring of the grazing environment.
In a specific embodiment, the data acquisition and analysis module of the vegetation coverage of the grazing area of the unmanned aerial vehicle system mainly comprises the functions of data acquisition, vegetation coverage analysis, and balance management of the livestock.
In particular, the balance of the grass and the livestock is an important detection index for accurate grazing management. Before grazing, unmanned in the system can patrol the pasture area around which grazing can be performed in advance, obtain vegetation image data and upload the vegetation image data to a system server.
Firstly, carrying out pretreatment such as quality evaluation screening, smooth noise reduction, image cutting and the like on collected image data, and then calculating regional vegetation coverage according to the treated image data; the pasture is divided into several areas which can be grazed in a common rotation mode according to the number of herds to be grazed, and vegetation coverage of each area is recorded (updated periodically).
In the grazing process, when the herd is in a relatively stable foraging state, the system schedules 1-2 unmanned aerial vehicles to fly above the herd based on a multi-machine cooperative algorithm, shoots vegetation image data of a grazing area through hovering a positive overlooking angle, uploads the vegetation image data to a system server, performs data preprocessing, calculates vegetation coverage of the current grazing area,
in addition, the system can estimate the grass feeding demand of the herd according to the actual grazing livestock quantity, and calculate the grass-livestock balance condition by combining the regional vegetation coverage, so as to make a decision whether the current grazing region needs to be changed or not: if the vegetation coverage condition of the current area is smaller than a grass-livestock balance threshold set by a system, driving the livestock group to an area with better vegetation coverage condition in the pasture; if the current grazing zone has not reached the herd balancing threshold, the herd need not be driven to rotate the grazing zone.
In a specific embodiment, the data acquisition and analysis module of the movable image of the herd of the unmanned aerial vehicle system mainly comprises the functions of data acquisition, individual identification of the herd, target detection and the like, and is used for assisting in automatic driving grazing.
When the herd is in a stable advancing or foraging state, the system schedules 1-2 unmanned aerial vehicles to fly above the herd based on a multi-machine collaborative algorithm, adjusts the flying height and the flying speed according to weather conditions and the general moving speed of the herd, shoots moving image data of the herd vertically and equidistantly through hovering and overlooking angles, and then uploads the moving image data to a system server.
The server performs preprocessing on the collected image data, including data screening and removing, image rotation, translation, overturning, scaling, cutting and the like to achieve data enhancement, and then divides the data set into a training set and a testing set.
And then, model training, verification, prediction and export are carried out on the Yolov8 target detection model based on the training set and the historical image data, so that individual identification of the herd, motion trail monitoring and prediction are realized.
The YOLOv8 target detection model introduces a new improved module, has great improvement in speed and performance, but in the practical application of an unmanned aerial vehicle system for identifying the herd individuals, the problems of unmanned aerial vehicle flight shake, poor consistency of data set shooting height, unbalanced sample quantity of different types and the like are unavoidable, and the training and performance of the model are influenced. In order to overcome the problems, the detection precision is further improved, and the data acquisition and analysis process of the unmanned aerial vehicle system is improved: firstly, in the data set collection stage, according to the types of the animal herd, more and more abundant individual characteristic data of livestock are collected as much as possible, wherein the characteristic data comprise images under different varieties, different ages and different illumination conditions so as to increase the generalization capability of the model; secondly, in the data preprocessing stage, data enhancement is realized through data screening and removing, and image rotation, translation, overturning, scaling, cutting and the like, so that the recognition capability of the model on individual targets is improved; thirdly, multi-scale training and reasoning can be carried out by using unmanned aerial vehicles to fly at different heights or angles in the model training and reasoning process to obtain image data of different scales so as to improve the recognition capability of the model on targets of different scales; fourthly, according to the training condition of the model, super parameters such as learning rate, regularization parameters and the like of the model are adjusted so as to find out model configuration which is more suitable for the current livestock type; fifthly, in order to solve the problem of unbalanced sample data of different categories, the categories are balanced by adopting methods such as over sampling, under sampling or category weighting, so that the model is prevented from neglecting the categories with fewer numbers; and sixthly, performing migration learning based on pre-trained model weights, taking the weights of the models for performing relevant task training on large-scale historical data as initial weights, freezing a part of weights of a network layer, replacing or adding a new output layer according to the requirements of actual target detection tasks, thawing a part of the network layer, allowing the network layer to fine-tune parameters on the new tasks, and training and optimizing the models by using newly acquired data sets so as to improve the training of the models and improve the performance of the models.
On the other hand, the embodiment of the invention discloses an unmanned aerial vehicle multi-machine collaborative natural pasture intelligent management system, which consists of 3-5 four-rotor unmanned aerial vehicles provided with sensors, cameras, stereo players and communication devices, readable storage media for storing computer program instructions and a processor for executing the computer programs, is used for multi-machine collaborative pasture and image acquisition functions, can complete the image information acquisition of the herd activities and pasture areas under the condition of not interfering with the normal activities of the herd, and drives the herd to a target place by playing sound through the unmanned aerial vehicles according to the requirements of pasture management. Under different conditions, unmanned in the system can bear different tasks such as grazing driving or image acquisition, and therefore, a corresponding path planning and multi-machine cooperative control algorithm is needed. The pasture vegetation coverage data that this system gathered can be used to real-time analysis grazing intensity, can drive the herd by the higher region of grazing intensity to lower region through unmanned aerial vehicle system to keep the grass and animal balanced, avoid excessively grazing to destroy grassland ecological function. The movable image data of the herd, which is collected by the system, can be used for analyzing the movable track of the herd and identifying the individual behaviors so as to carry out accurate grazing management.
For the system device disclosed in the embodiment, since the system device corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. An intelligent management method for a natural pasture with multiple unmanned aerial vehicles in cooperation is characterized by comprising the following steps:
acquiring pasture environment information through an unmanned aerial vehicle group;
the unmanned aerial vehicle group automatically plans a global grazing path according to the pasture environment information, the obtained user-defined grazing time and the starting point and the ending point, and automatically drives grazing;
in the automatic driving grazing process based on the global grazing path, acquiring and analyzing movable images of the herd, planning a local grazing path, and assisting in automatic driving grazing;
and in the automatic driving grazing process based on the global grazing path, collecting and analyzing vegetation coverage image data of the grazing area to assist in automatic driving grazing.
2. The intelligent management method for natural pasture coordinated with multiple unmanned aerial vehicles according to claim 1, wherein the obtaining pasture environment information by the unmanned aerial vehicle group comprises the following steps:
acquiring pasture environment image data through a binocular depth camera carried by the unmanned aerial vehicle;
preprocessing the pasture environment image data, and carrying out distortion correction and stereo correction based on a stereo matching method to obtain three-dimensional dense point cloud data;
constructing an octree map based on the three-dimensional dense point cloud data, dynamically adjusting the precision and the resolution, screening out effective three-dimensional dense point cloud data, and constructing a pasture area environment map;
dividing the pasture area environment map into a plurality of areas capable of being alternately grazed, marking obstacles affecting normal flight of the unmanned aerial vehicle group in each area, and storing vegetation coverage conditions of each area to a cloud server to obtain pasture environment information.
3. The intelligent management method for natural pastures by unmanned aerial vehicle multi-machine cooperation according to claim 2, wherein the vegetation coverage condition of each area is stored in a cloud server, and the method comprises the following steps: the method comprises the steps of obtaining pasture vegetation image data through an unmanned aerial vehicle group, carrying out quality evaluation screening, smooth noise reduction and image cutting on the pasture vegetation image data, calculating regional vegetation coverage according to the processed pasture vegetation image data, and storing the regional vegetation coverage in a cloud server.
4. The intelligent management method for the natural pasture coordinated with multiple unmanned aerial vehicles according to claim 1, wherein the unmanned aerial vehicle group automatically plans a global pasture path according to the pasture environment information, the obtained user-defined pasture time and a starting point end point to automatically drive pasture, and the method comprises the following steps:
acquiring a user-defined grazing starting point and end point region and grazing time;
calling a global path planner based on an A algorithm to plan a global grazing path according to pasture environment information and a starting point and end point region, wherein the method comprises the following steps:
selecting Manhattan distance between two points of the path as the estimation of the distance between the two points;
the method comprises the steps of acquiring elevation data of terrain through a sensor carried by an unmanned aerial vehicle, establishing a digital elevation model, and adjusting a cost function of a path according to the digital elevation model:
the actual cost function from the starting node to node n is formulated as follows:
G(n)＝G(parent)+Cost(n,parent)；
where the parent node parent is the previous point to node n, G (parent) is the actual Cost from the start point to the parent node, and Cost (n, parent) is the extra Cost from the parent node to node n;
the heuristic estimation cost function H considers the path cost from the node n to the target node gold, and the formula is as follows:
H(n)＝k 1 ×Distance(n,parent)+k 2 ×|Altitude(n)-Altitude(goal)|；
the path cost function is:
F(n)＝G(n)+H(n)
where G (n) and H (n) are the actual cost and the heuristically estimated cost function, the computed path cost function will be used to select the next node to be expanded and generate the global grazing path.
5. The intelligent management method for natural pasture coordinated with multiple unmanned aerial vehicles according to claim 4, wherein the calculation formula of the extra Cost (n, parent) from the parent node to the node n is as follows:
Cost(n,parent)＝k 1 ×Distance(n,parent)+k 2 ×|Altitude(n)-Altitude(parent)|+k 3 ×Speed_Adjusted_Cost(n)+k 4 ×Obstacle_Cost(n)；
wherein Distance (n, parent) is the Manhattan Distance between two points; altitude (n) is the elevation of node n; speed_adjusted_cost (n) is a Cost related to unmanned aerial vehicle flight Speed, including unmanned aerial vehicle flight Speed and herd movement Speed; obstacle_cost (n) is the Cost associated with the obstacles around node n, and is positively related to the density or distance of the obstacles around the path, k 1 、k 2 、k 3 、k 4 Is the weight.
6. The intelligent management method for natural pasture coordinated with multiple unmanned aerial vehicles according to claim 1, wherein in the process of automatically driving grazing based on a global grazing path, the method for acquiring and analyzing the moving images of the herd, planning a local grazing path and assisting the automatic driving grazing comprises the following steps: in the automatic driving grazing process based on the global grazing path, a movable image of the herd is acquired and analyzed to obtain the position and the motion state of the herd in the actual grazing process, a local path planner based on an improved artificial potential field method is called to plan out a local path of the unmanned aerial vehicle grazing according to the position and the motion state of the herd in the actual grazing process, and the herd is driven to a target grazing area through a stereo player carried by the unmanned aerial vehicle, so that auxiliary automatic driving grazing is realized.
7. The intelligent management method for natural pastures by unmanned aerial vehicle multi-machine cooperation according to claim 6, wherein the steps of collecting and analyzing the movable images of the herd to obtain the position and the movement state of the herd in the actual grazing process comprise:
collecting movable image data of the livestock group through the unmanned aerial vehicle group;
preprocessing the acquired movable image data of the livestock group, including data screening and removing, and performing image rotation, translation, turnover, scaling and cutting processing to generate a data set;
model training, verification, prediction and derivation are carried out on the YOLOv8 target detection model based on the data set, individual identification of the herd and motion trail monitoring and prediction are achieved, and the position and motion state of the herd in the actual grazing process are obtained.
8. The method for intelligent management of natural pasture in cooperation with multiple unmanned aerial vehicles according to claim 6, wherein the step of calling the local path planner based on the improved artificial potential field method to plan the local path of the unmanned aerial vehicle pasture comprises the following steps: the rejection field is expressed using a gaussian function for pointing away from the herd, as follows:
wherein F is repulsion Is the vector of the repulsive force field; n is the number of herd individuals; k is a coefficient for adjusting the intensity of the repulsive force; alpha is a parameter for adjusting the attenuation of the repulsive force; d, d i Is the distance from the unmanned aerial vehicle to the ith herd individual; (x) i ,y i ) Is the location of the ith herd individual; (x, y) is the current position of the drone.
9. The method for intelligent management of natural pasture in cooperation with multiple unmanned aerial vehicles according to claim 6, wherein the step of calling the local path planner based on the improved artificial potential field method to plan the local path of the unmanned aerial vehicle pasture comprises the following steps: the gravity field is expressed by using a Gaussian function, and the individual formula of the herd closest to the unmanned plane is as follows:
wherein F is attraction Is the vector of the gravitational field; n is the number of herd individuals; w is a coefficient for adjusting the strength of attraction; beta is a parameter for adjusting the attenuation of the gravitational strength; d, d i Is the distance from the unmanned aerial vehicle to the ith herd individual; (x) i ,y i ) Is the location of the ith herd individual; (x, y) is the current position of the drone.
10. An unmanned aerial vehicle multi-machine collaborative natural rangeland intelligent management system using the unmanned aerial vehicle multi-machine collaborative natural rangeland intelligent management method of any of claims 1-9, comprising: the unmanned aerial vehicle group, the readable storage medium storing the computer program, a processor executing the computer program and a cloud server, wherein the unmanned aerial vehicle group consists of four-rotor unmanned aerial vehicles provided with sensors, binocular depth cameras, stereo players and communication devices, and the computer program realizes the steps of the unmanned aerial vehicle multi-machine collaborative natural pasture intelligent management method according to any one of claims 1-9 when being executed by the processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311494514.2A CN117434965A (en) | 2023-11-10 | 2023-11-10 | Unmanned aerial vehicle multi-machine collaborative natural pasture intelligent management method and management system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311494514.2A CN117434965A (en) | 2023-11-10 | 2023-11-10 | Unmanned aerial vehicle multi-machine collaborative natural pasture intelligent management method and management system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117434965A true CN117434965A (en) | 2024-01-23 |
Family
ID=89547850
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311494514.2A Pending CN117434965A (en) | 2023-11-10 | 2023-11-10 | Unmanned aerial vehicle multi-machine collaborative natural pasture intelligent management method and management system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117434965A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117311381A (en) * | 2023-09-20 | 2023-12-29 | 中国农业大学 | Multi-unmanned aerial vehicle intelligent inspection system and method based on vehicle-mounted mobile nest |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107219860A (en) * | 2017-07-31 | 2017-09-29 | 内蒙古智牧溯源技术开发有限公司 | A kind of unmanned plane rang management system and method |
| CN107229289A (en) * | 2017-07-31 | 2017-10-03 | 内蒙古智牧溯源技术开发有限公司 | A kind of unmanned plane grazing management system |
| CN107544553A (en) * | 2017-10-11 | 2018-01-05 | 湖北工业大学 | A kind of Path Planning for UAV based on hybrid ant colony |
| CN109933083A (en) * | 2017-12-15 | 2019-06-25 | 翔升（上海）电子技术有限公司 | Grazing method, device and system based on unmanned plane |
| CN109960272A (en) * | 2017-12-22 | 2019-07-02 | 翔升（上海）电子技术有限公司 | Grazing method and system based on unmanned plane |
| KR20190098804A (en) * | 2018-01-31 | 2019-08-23 | 주식회사 에디테크놀로지 | System for controlling position of grazing animals using drones |
| CN116182838A (en) * | 2023-03-27 | 2023-05-30 | 江南大学 | Plant protection unmanned aerial vehicle global path optimization method based on point cloud map |
-
2023
- 2023-11-10 CN CN202311494514.2A patent/CN117434965A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107219860A (en) * | 2017-07-31 | 2017-09-29 | 内蒙古智牧溯源技术开发有限公司 | A kind of unmanned plane rang management system and method |
| CN107229289A (en) * | 2017-07-31 | 2017-10-03 | 内蒙古智牧溯源技术开发有限公司 | A kind of unmanned plane grazing management system |
| CN107544553A (en) * | 2017-10-11 | 2018-01-05 | 湖北工业大学 | A kind of Path Planning for UAV based on hybrid ant colony |
| CN109933083A (en) * | 2017-12-15 | 2019-06-25 | 翔升（上海）电子技术有限公司 | Grazing method, device and system based on unmanned plane |
| CN109960272A (en) * | 2017-12-22 | 2019-07-02 | 翔升（上海）电子技术有限公司 | Grazing method and system based on unmanned plane |
| KR20190098804A (en) * | 2018-01-31 | 2019-08-23 | 주식회사 에디테크놀로지 | System for controlling position of grazing animals using drones |
| CN116182838A (en) * | 2023-03-27 | 2023-05-30 | 江南大学 | Plant protection unmanned aerial vehicle global path optimization method based on point cloud map |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117311381A (en) * | 2023-09-20 | 2023-12-29 | 中国农业大学 | Multi-unmanned aerial vehicle intelligent inspection system and method based on vehicle-mounted mobile nest |
| CN117311381B (en) * | 2023-09-20 | 2024-03-26 | 中国农业大学 | Multi-unmanned aerial vehicle intelligent inspection system and method based on vehicle-mounted mobile nest |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Namani et al. | Smart agriculture based on IoT and cloud computing | |
| CN110874578B (en) | Unmanned aerial vehicle visual angle vehicle recognition tracking method based on reinforcement learning | |
| CN113807017B (en) | Method for determining fish preference habitat and terminal equipment | |
| CN108830373A (en) | The modeling method that the extensive intelligent group of imitative starling cluster flight independently cooperates with | |
| CN117434965A (en) | Unmanned aerial vehicle multi-machine collaborative natural pasture intelligent management method and management system | |
| CN112684807A (en) | Unmanned aerial vehicle cluster three-dimensional formation method | |
| CN112469050B (en) | WSN three-dimensional coverage enhancement method based on improved wolf optimizer | |
| CN113597874B (en) | Weeding robot and weeding path planning method, device and medium thereof | |
| CN113421345B (en) | Bionic robot fish cluster navigation simulation method based on deep reinforcement learning technology | |
| Hartanto et al. | Intelligent unmanned aerial vehicle for agriculture and agroindustry | |
| Ratnayake et al. | Spatial monitoring and insect behavioural analysis using computer vision for precision pollination | |
| Brace et al. | Using collision cones to assess biological deconfliction methods | |
| Zhilenkov et al. | The use of convolution artificial neural networks for drones autonomous trajectory planning | |
| Puente-Castro et al. | Q-learning based system for path planning with unmanned aerial vehicles swarms in obstacle environments | |
| Kassim et al. | Small object bird detection in infrared drone videos using mask R-CNN deep learning | |
| Li | Some problems of deployment and navigation of civilian aerial drones | |
| CN116301042A (en) | Unmanned aerial vehicle group autonomous control method based on VGG16 and virtual game | |
| CN116400728A (en) | Unmanned plane path planning method based on depth deterministic strategy gradient algorithm | |
| Haque et al. | Geometric foraging strategies in multi-agent systems based on biological models | |
| Xue et al. | Monocular vision obstacle avoidance UAV: a deep reinforcement learning method | |
| Chen et al. | Novel intelligent grazing strategy based on remote sensing, herd perception and UAVs monitoring | |
| CN117556979B (en) | Unmanned plane platform and load integrated design method based on group intelligent search | |
| Zhang et al. | Position Planning for Monitoring Shepherd-UAVs Using the Herd’s Location Information | |
| CN113589810B (en) | Dynamic autonomous obstacle avoidance movement method and device for intelligent body, server and storage medium | |
| Li | Deployment and navigation of aerial drones for sensing and interacting applications |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |