Multi-unmanned aerial vehicle cooperative positioning method for economic area under forest

Abstract

The invention provides an under-forest economic area multi-unmanned aerial vehicle cooperative positioning method, and belongs to the technical field of positioning navigation application. The method comprises the following steps: acquiring an all-sky photo of an under-forest point location, and calculating the crown opening degree of the point location; the hovering positioning time of the unmanned aerial vehicle is set according to the crown opening degree, and under-forest hovering positioning of the base station unmanned aerial vehicle is carried out by using satellite signals; positioning the under-forest sensing unmanned aerial vehicle cluster based on the position information of the base station unmanned aerial vehicle; on the basis of multi-unmanned-aerial-vehicle cooperation of the base station unmanned aerial vehicle and the sensing unmanned aerial vehicle cluster, under-forest positioning signals are explored, and multi-unmanned-aerial-vehicle cooperative under-forest positioning is achieved According to the method, efficient and accurate positioning of the under-forest economic area can be achieved through cooperation of satellite signals and the multiple unmanned aerial vehicles, the position information service of vertical integration of the canopy is promoted, mechanical and automatic support is provided for under-forest planting and breeding, and intensive and efficient development of under-forest economy is assisted.

Description

Technical Field

[0001] The present invention relates to the field of positioning and navigation application technology, and in particular to a method for collaborative positioning of multiple unmanned aerial vehicles in a forest economic zone. Background Art

[0002] With the continuous advancement of agricultural modernization, the application of global satellite navigation systems (GNSS) such as the Beidou system in the agricultural field is becoming more and more extensive, including automated navigation driving, autonomous operation of agricultural machinery, and drone plant protection. These technologies have achieved precise navigation and operation in open fields, greatly improving agricultural production efficiency. However, in forest economic areas, especially in hilly and mountainous areas, the application of traditional large wheeled agricultural machinery is greatly restricted due to the complex terrain; in addition, the ground space of forest economy may be irregularly planted or raised due to terrain restrictions, which further increases the difficulty of mechanical operation. Therefore, in forest economic areas with complex terrain, more intelligent micro-agricultural machinery is needed to meet the special needs of forest economy. However, in forest economic areas, due to the complex terrain and canopy occlusion, traditional GNSS signals are difficult to penetrate, resulting in challenges in automated positioning and navigation services, which also restricts the mechanization and automation of forest planting and forest breeding.

[0003] In recent years, the development of drone technology has provided new solutions for forest economic areas, especially the collaborative perception and positioning mapping of drone swarms, which can effectively improve the accuracy of positioning. However, in forest spaces, the collaborative positioning and navigation services of drone swarms still face technical bottlenecks. The main problems include:

[0004] (1) GNSS signal blocking: The canopy blocks the GNSS signal, making it difficult for the drone to obtain stable positioning information.

[0005] (2) Lack of ground control points: The forest economic area lacks field correction control points with absolute position information, which limits the precise application of remote sensing data.

[0006] (3) Insufficient collaborative positioning technology: Existing technologies have not yet formed a mature collaborative positioning solution for drone clusters, especially in forest environments.

[0007] Therefore, it is necessary to improve the existing positioning method of drone clusters to overcome the shortcomings of the existing technology. Summary of the invention

[0008] In order to overcome the problems existing in the related technologies, the purpose of the present invention is to provide a method for collaborative positioning of multiple UAVs in the forest economic zone. The method can utilize satellite signals and the collaboration of multiple UAVs to achieve efficient and accurate positioning of the forest economic zone, promote integrated location information services above and below the forest canopy, provide mechanized and automated support for forest planting and breeding, and help the intensive and efficient development of the forest economy.

[0009] A method for collaborative positioning of multiple UAVs in an understory economic zone, comprising:

[0010] Obtain full-sky photos of forest understory points and calculate the canopy openness of the points;

[0011] Set the drone hovering positioning time according to the canopy width, and use satellite signals to perform base station drone hovering positioning under the forest;

[0012] Based on the location information of the base station drone, the understory sensing drone cluster is positioned;

[0013] Based on the collaboration of multiple UAVs including base station UAVs and perception UAV clusters, we explore the understory positioning signals and realize the collaborative understory positioning of multiple UAVs.

[0014] In a preferred technical solution of the present invention, the step of obtaining a full sky photo of a point under the forest and calculating the canopy openness of the point includes:

[0015] Using a photosensitive element to trigger a camera module to take a photo, obtaining a full-sky photo of a forest understory point that can receive a satellite signal; wherein the satellite signal is a Beidou satellite signal;

[0016] Calculate the canopy openness of a point based on a deep learning algorithm.

[0017] In a preferred technical solution of the present invention, the method of using a photosensitive element to trigger a camera module to take a photo, and obtaining a full-sky photo of a forest point that can receive satellite signals, includes:

[0018] The threshold of sunlight brightness change per unit time is selected as the criterion for selecting candidate locations for understory positioning;

[0019] Use photosensitive elements to monitor the light intensity on the top of the drone in real time. When the photosensitive element detects that the light intensity exceeds the set threshold, the circular camera module is triggered to start taking pictures.

[0020] During the photo shooting process, use a wide-angle fisheye lens to shoot photos with a hemispherical field of view greater than 150°;

[0021] Correct the captured fisheye image to eliminate the distortion caused by the fisheye lens.

[0022] Record and store corrected full-sky images.

[0023] Among them, the correction processing of the captured fisheye image includes "normal correction, longitude and latitude expansion correction, surround expansion correction, arbitrary perspective correction and multi-perspective framing correction" and other methods to correct the fisheye image.

[0024] In a preferred technical solution of the present invention, the threshold value of sunlight brightness change per unit time is selected as the standard for selecting candidate locations for understory positioning, including:

[0025] The sunlight brightness is calculated using the reverse ray tracing method; specifically, the light starts from the camera, and the light entering the camera is traced back to calculate the brightness of each point in the camera's field of view; by assuming that the surface of the object is an ideal mirror reflection, the reflected light is traced to the light source, and the scattering distribution during the light transmission process is recorded, thereby deriving the basic formula for the brightness calculation equation:

[0026]

[0027] Among them, ω i is the spatial solid angle of the i-th ray, φ is the angular range of the spatial solid angle, L ei (P,ω0,λ) is the incident spectral radiance, θ i is the angle between the incident light ray and the normal element;

[0028] f γ (P,ω i ,ω0,λ) is the spectral reflectivity of the object surface to incident light in any direction in any direction. The specific expression is as follows:

[0029]

[0030] Using the basic formula of the brightness calculation equation, calculate the spectral radiance data of each point on the camera's viewing plane, and calculate the camera's radiance value and the brightness value perceived by the camera;

[0031] The camera's radiation brightness value is calculated as follows:

[0032]

[0033] The brightness value perceived by the camera is calculated as follows:

[0034]

[0035] Among them, K m It is expressed as the maximum spectral visual efficacy under bright vision conditions, and V(λ) represents how many lumens of single-spectrum luminous flux can be converted into per watt of single-spectrum radiant flux;

[0036] The brightness information in the forest environment is obtained according to the camera's radiation brightness value and the brightness value perceived by the camera, so as to calculate the sunlight brightness per unit time and the threshold value of sunlight brightness change per unit time;

[0037] When the light intensity at the top of the drone increases above this threshold, the point is considered a candidate point for understory positioning.

[0038] In a preferred technical solution of the present invention, the method of calculating the canopy openness of a point based on a deep learning algorithm includes:

[0039] The canopy openness refers to the proportion of the sky sphere that is not blocked by tree branches when looking up from a point in the forest.

[0040] The calculation of canopy openness is achieved by calculating canopy closure, where canopy closure is the ratio of the projected area of ​​the canopy to the forest area;

[0041] The relationship between canopy openness and canopy closure is as follows:

[0042] Canopy openness = 1-canopy closure;

[0043] The calculation process of canopy closure is as follows:

[0044] Obtain the full sky fisheye image of the forest understory point, and set the actual area represented by each pixel under different resolution conditions to f(d), that is:

[0045] Pa=f(d)

[0046] Where Pa is the actual area represented by the pixel in the image; d is the resolution of the scanner;

[0047] Calculate the positive area S of the leaf in the acquired image. The specific calculation formula is as follows:

[0048]

[0049] in, The integral of the number of pixels contained in the leaf image contour in the acquired image;

[0050] The canopy closure is calculated using the leaf orthophoto area S and the total area of ​​the acquired image;

[0051] or,

[0052] The calculation process of canopy closure is as follows:

[0053] Obtain all-sky fisheye images of understory locations;

[0054] digitizing the acquired image, and calculating the number of pixels contained in the plant leaf image from the digitized image;

[0055] The canopy closure is calculated using the number of pixels contained in the plant leaf image and the number of pixels contained in the entire image.

[0056] In a preferred technical solution of the present invention, the method of setting the drone hovering positioning time according to the canopy width and using satellite signals to perform under-forest hovering positioning of the base station drone includes:

[0057] Using the intelligent perception of photosensitive elements, multiple candidate locations for understory positioning were selected, and a 5-minute drone hovering flight positioning experiment was conducted at the candidate locations.

[0058] The canopy openness of the points where the drone can hover and locate within 5 minutes is counted, and based on the statistics, the minimum threshold of canopy openness for drone understory positioning that can obtain satellite signal support is determined;

[0059] The candidate points under the forest where the canopy openness is greater than the minimum threshold are located, and the drone hovering flight positioning experiments with durations of 1min, 30s, 20s, 10s, 5s, and 3s are conducted;

[0060] Record and analyze the situation of the drone achieving a fixed positioning solution in each flight time period under different canopy openness;

[0061] Based on the experimental data, the impact of different hovering times on positioning accuracy is evaluated, and the optimal solution for the hovering time and positioning accuracy of the drone is obtained;

[0062] According to the best solution, satellite signals are used to hover and locate the base station drone under the forest.

[0063] In a preferred technical solution of the present invention, the positioning of candidate points under the forest where the canopy openness is greater than the minimum threshold is performed by conducting a drone hovering flight positioning experiment for a duration of 1 minute, 30 seconds, 20 seconds, 10 seconds, 5 seconds, or 3 seconds, including:

[0064] Control the drone to hover at the candidate point for a designed duration and locate it through the received satellite signals;

[0065] In each experiment, record whether the drone can successfully obtain a fixed positioning solution and the specific time required for positioning;

[0066] Analyze the impact of different hovering times on the positioning success rate, positioning speed and positioning accuracy to evaluate the positioning performance of the UAV under different hovering times;

[0067] Get the shortest hovering time required by the drone while ensuring positioning accuracy.

[0068] In a preferred technical solution of the present invention, the positioning of the understory sensing drone cluster based on the location information of the base station drone includes:

[0069] By measuring the distance or angle data between the base station drone and the perception drone, the position of the perception drone is determined using a geometric algorithm;

[0070] Using the communication and measurement equipment between the base station drone and the perception drone, the distance or angle data between the base station drone and each perception drone is collected in real time;

[0071] Use the distance resection method or free station method to calculate the flight trajectory and state parameters of each sensing drone in real time based on the collected distance or angle data.

[0072] In a preferred technical solution of the present invention, the multi-UAV collaboration based on the base station UAV and the perception UAV cluster explores the understory positioning signal and realizes the multi-UAV collaborative understory positioning, including:

[0073] In the process of multi-UAV coordinated forest positioning, a dynamic rotation mechanism of the "leader" base station UAV group is established to dynamically alternate the identities of the base station UAV and the perception UAV;

[0074] Among them, after the perception drone finds the candidate positioning point area under Helin, it can be converted into a base station drone, and perform rapid static measurement and positioning with other base station drones, autonomously exploring and collaborating with each other;

[0075] The base station drone adjusts its altitude up and down when hovering to obtain the best satellite positioning signal, and the sensing drone flies at a fixed reference altitude.

[0076] In a preferred technical solution of the present invention, the construction of a dynamic rotation mechanism of the "leader" base station drone group, dynamically alternating the identities of the base station drone and the perception drone, includes:

[0077] In the process of multiple drones coordinating under-forest positioning, the “leader” base station drone group uses edge connection to form a network, forming a strip network composed of triangles or geodetic quadrilaterals;

[0078] During the dynamic rotation of the base station drone group, at least three base station drones are ensured to be in a hovering positioning state during the station relocation process;

[0079] Each time the station is moved, the base station drone moves forward to explore satellite positioning signals based on its own navigation trajectory and real-time positioning results.

[0080] The beneficial effects of the present invention are:

[0081] The present invention provides a method for collaborative positioning of multiple UAVs in an understory economic zone, the method comprising: obtaining a full sky photo of an understory point and calculating the canopy openness of the point; setting the hovering positioning time of the UAV according to the canopy openness, and using satellite signals to perform understory hovering positioning of the base station UAV; positioning the understory sensing UAV cluster based on the location information of the base station UAV; and exploring the understory positioning signal based on the collaboration of multiple UAVs of the base station UAV and the sensing UAV cluster, so as to achieve collaborative understory positioning of multiple UAVs. The method adjusts the hovering time in real time according to different canopy openness conditions through the hovering flight mode of multiple UAVs, thereby effectively improving the positioning efficiency in a complex understory environment. In addition, the method can use GNSS satellite signals such as Beidou to assist in positioning, and combine algorithms to quickly calculate the canopy openness, optimize the UAV hovering time and positioning strategy, which can improve the final positioning accuracy; in addition, the method uses the collaborative work of multiple UAVs of the base station UAV and the sensing UAV cluster for positioning, which improves the fault tolerance and self-healing ability of the overall system. In summary, this method can achieve efficient and accurate positioning of the forest economic zone, promote integrated location information services above and below the canopy, provide mechanized and automated support for forest planting and breeding, and help the intensive and efficient development of the forest economy. This method has a wide range of applications. In addition to being applicable to forest economic zones, it can also be widely used in UAV collaborative positioning tasks in other complex environments, such as mountainous areas, forests, and urban canyons. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 is a flow chart of a method for collaborative positioning of multiple UAVs in an understory economic zone provided in an embodiment of the present invention;

[0083] Figure 2 It is a flow chart provided in an embodiment of the present invention and calculating the canopy openness of a point;

[0084] Figure 3 is a flow chart of how to obtain a full sky photo of a forest understory point that can receive satellite signals, provided in an embodiment of the present invention;

[0085] Figure 4 It is a schematic diagram of obtaining a full-sky photo of a forest understory point that can receive satellite signals such as Beidou provided in an embodiment of the present invention;

[0086] Figure 5 Schematic diagram of a hovering flight positioning experiment of a drone at different forest locations provided in an embodiment of the present invention;

[0087] Figure 6 It is a schematic diagram of positioning a cluster of understory sensing drones provided in an embodiment of the present invention;

[0088] Figure 7It is a schematic diagram of how to perform multi-UAV collaborative understory positioning based on multi-UAV collaboration provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0089] The preferred embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the preferred embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to fully convey the scope of the present invention to those skilled in the art.

[0090] In the existing technology, the collaborative positioning and navigation services of drone swarms in the forest space still face technical bottlenecks. The main problems include the difficulty for drones to obtain stable positioning information due to obstruction by the forest canopy, the lack of field correction control points with absolute position information, and the lack of mature collaborative positioning solutions for drone swarms.

[0091] Based on this, the present application provides a method for collaborative positioning of multiple UAVs in forest economic zones.

[0092] See also Figure 1-Figure 7 The present application provides a method for collaborative positioning of multiple UAVs in an understory economic zone, including:

[0093] S100, obtain a full sky photo of a point under the forest, and calculate the canopy openness of the point; canopy openness refers to the proportion of the sky sphere that is not blocked by tree branches when looking up from a point in the forest. It is the central projection of the point scale and is consistent with the hemispherical field of view environment for GNSS positioning;

[0094] Specifically, obtaining a full sky photo of a point under the forest and calculating the canopy openness of the point includes:

[0095] S110, using a photosensitive element to trigger a camera module to take photos, and obtaining full-sky photos of forest locations that can receive satellite signals; wherein the satellite signal is a Beidou satellite signal; the camera module and the photosensitive element are both mounted on the drone, that is, the drone has its own camera module and photosensitive element.

[0096] Furthermore, the method of using a photosensitive element to trigger a camera module to take a photo, and obtaining a full-sky photo of a forest understory point that can receive satellite signals, includes:

[0097] S1101, selecting a threshold value of sunlight brightness change per unit time as a standard for selecting candidate locations for understory positioning;

[0098] S1102. Use a photosensitive element to monitor the light intensity on the top of the drone in real time. When the photosensitive element detects that the light intensity exceeds a set threshold, the circular camera module is triggered to start taking pictures. When the light intensity exceeds the set threshold, it indicates that the current environment is suitable for taking pictures.

[0099] S1103. During the photo-taking process, a wide-angle fisheye lens is used to take photos of a hemispherical field of view greater than 150°. This is done to obtain the largest possible sky field of view to provide more information for subsequent positioning analysis.

[0100] S1104, correcting the captured fisheye image to eliminate distortion caused by shooting with a fisheye lens; wherein, the correcting of the captured fisheye image includes correcting the fisheye image by methods such as “normal correction, longitude and latitude expansion correction, surround expansion correction, arbitrary viewing angle correction and multi-viewing angle framing correction”.

[0101] S1105: Record and store the corrected full sky image.

[0102] Through the above steps, we can ensure that high-quality full-sky photos are captured under suitable lighting conditions. These photos will provide key data support for subsequent precise positioning using satellite signals.

[0103] S120. Calculate the canopy openness of a point based on a deep learning algorithm; canopy openness refers to the proportion of the sky sphere that is not blocked by tree branches when looking up from a point in the forest;

[0104] In this step, the canopy openness is calculated using the processed full sky image. The specific steps include:

[0105] The method for calculating canopy openness is as follows:

[0106] Method 1: Calculation of canopy openness is achieved by calculating canopy closure, where canopy closure is the ratio of the canopy projection area to the forest area;

[0107] The relationship between canopy openness and canopy closure is as follows:

[0108] Canopy openness = 1-canopy closure;

[0109] The calculation process of canopy closure is as follows:

[0110] Obtain the full sky fisheye image of the forest understory point, and set the actual area represented by each pixel under different resolution conditions to f(d), that is:

[0111] Pa=f(d)

[0112] Where Pa is the actual area represented by the pixel in the image; d is the resolution of the scanner;

[0113] Calculate the positive area S of the leaf in the acquired image. The specific calculation formula is as follows:

[0114]

[0115] in, The integral of the number of pixels contained in the leaf image contour in the acquired image;

[0116] The canopy closure is calculated using the orthophoto area S of the leaves and the total area of ​​the acquired image; the unit of the orthophoto area S may be square meters.

[0117] Specifically, the canopy closure is the ratio of the leaf orthogonal area S to the total area of ​​the acquired image. This method indirectly obtains the canopy openness by calculating the canopy closure, which improves the accuracy of canopy openness measurement. This method uses high-resolution all-sky fisheye images and combines advanced image processing technology to accurately identify and calculate the projection area of ​​the canopy, thereby obtaining a more accurate canopy closure.

[0118] In practical applications, the use of deep learning and image processing algorithms to automatically calculate forest canopy closure greatly reduces the workload of manual measurement and calculation and improves efficiency. This method can process a large amount of image data in a relatively short period of time, providing rapid feedback for the management and monitoring of forest economic zones.

[0119] Method 2: The calculation process of canopy closure is as follows:

[0120] Obtain all-sky fisheye images of understory locations;

[0121] digitizing the acquired image, and calculating the number of pixels contained in the plant leaf image from the digitized image;

[0122] The canopy closure is calculated using the number of pixels contained in the plant leaf image and the number of pixels contained in the entire image.

[0123] Specifically, in the all-sky fisheye images of the forest understory points, there is a big difference between the leaf color and the sky blank, so the computer can be used to reclassify the pixels of the all-sky images into two categories: blank and main color of plant leaves. The number of pixels in the two categories is then counted in the classified pixel attribute table, and the canopy closure and canopy openness can be obtained by calculation according to the formula.

[0124] The following is a detailed process of how to use the convolutional neural network algorithm to calculate the canopy openness:

[0125] Data collection and preprocessing: First, a large number of all-sky fisheye images of forest understory points need to be collected. The images are preprocessed, including resizing, normalization, and contrast enhancement, to improve the training efficiency and accuracy of the model.

[0126] Image annotation: Manually annotate the images and mark the canopy and sky areas in the images. These annotated data will be used to train the CNN model to identify the canopy and sky in the images.

[0127] Design CNN architecture: Design a CNN architecture for image segmentation, which includes multiple convolutional layers, pooling layers, activation functions, and fully connected layers. The convolutional layer is used to extract image features, the pooling layer is used to reduce the spatial dimension of the features and reduce the amount of calculation, the activation function (such as ReLU) is used to introduce nonlinearity, and the fully connected layer is used for the final classification.

[0128] Model training: Use annotated image datasets to train the CNN model. During the training process, the model learns image features through forward propagation and adjusts the network weights through backpropagation to minimize the loss function (such as cross entropy loss).

[0129] Image segmentation: The trained CNN model can perform real-time segmentation on new full-sky images, separating the canopy and sky areas in the image. The segmentation map output by the model can be a binary image, in which the canopy and sky are marked with different pixel values.

[0130] Calculate canopy closure: Based on the segmentation results, calculate the ratio of the number of pixels in the canopy area to the number of pixels in the entire image, which is the canopy closure. Use the formula: canopy closure = (number of pixels in the canopy area) / (total number of pixels in the image).

[0131] Calculate the canopy openness: Calculate the canopy openness based on the canopy closure using the formula: canopy openness = 1 - canopy closure.

[0132] Result verification and optimization:

[0133] The output of the model was verified to ensure that the calculation results of canopy openness were accurate and reliable.

[0134] The model is adjusted and optimized based on the verification results to improve the accuracy and efficiency of the calculation.

[0135] In a more specific embodiment, the threshold value of sunlight brightness change per unit time is selected as the criterion for selecting candidate locations for understory positioning, including:

[0136] The sunlight brightness is calculated using the reverse ray tracing method; specifically, the light starts from the camera, and the light entering the camera is traced back to calculate the brightness of each point in the camera's field of view; by assuming that the surface of the object is an ideal mirror reflection, the reflected light is traced to the light source, and the scattering distribution during the light transmission process is recorded, thereby deriving the basic formula for the brightness calculation equation:

[0137]

[0138] Among them, ω i is the spatial solid angle of the i-th ray, φ is the angular range of the spatial solid angle, L ei (P,ω0,λ) is the incident spectral radiance, θ i is the angle between the incident light ray and the normal element;

[0139] f γ (P,ω i ,ω0,λ) is the spectral reflectivity of the object surface to incident light in any direction in any direction. The specific expression is as follows:

[0140]

[0141] Using the basic formula of the brightness calculation equation, calculate the spectral radiance data of each point on the camera's viewing plane, and calculate the camera's radiance value and the brightness value perceived by the camera;

[0142] The camera's radiation brightness value is calculated as follows:

[0143]

[0144] The brightness value perceived by the camera is calculated as follows:

[0145]

[0146] Among them, K m It is expressed as the maximum spectral visual efficacy under bright vision conditions, and V(λ) represents how many lumens of single-spectrum luminous flux can be converted into per watt of single-spectrum radiant flux;

[0147] The brightness information in the forest environment is obtained according to the camera's radiation brightness value and the brightness value perceived by the camera, so as to calculate the sunlight brightness per unit time and the threshold value of sunlight brightness change per unit time;

[0148] When the light intensity at the top of the drone increases above this threshold, the point is considered a candidate point for understory positioning.

[0149] S200. Set the drone hovering positioning time according to the canopy openness, and use satellite signals to hover and position the base station drone under the forest; in areas with higher openness, the hovering time can be shorter because the satellite signal reception quality is good; in areas with lower openness, the hovering time needs to be increased to ensure accurate reception of sufficient satellite signals for positioning. Within the preset hovering positioning time, the drone flies over the target area and hovers. The drone receives signals from satellites through its satellite navigation system receiver and calculates its precise position. During hovering, the drone should remain stable to reduce position errors.

[0150] Specifically, the method of setting the drone hovering positioning time according to the canopy openness and using satellite signals to perform under-forest hovering positioning of the base station drone includes:

[0151] Using the intelligent perception of photosensitive elements, multiple candidate locations for understory positioning are screened out, and a 5-minute drone hovering flight positioning experiment is conducted at the candidate locations. When the drone flies over the forest, the light changes detected by the photosensitive elements help identify areas that may receive satellite signals, thereby screening out multiple candidate locations for understory positioning.

[0152] The canopy openness of the points where the drone can hover and locate within 5 minutes is counted. Based on the statistics, the minimum threshold of the canopy openness for drone understory positioning supported by satellite signals is determined; specifically, based on the statistics, the minimum threshold of the canopy openness for drone understory positioning supported by satellite signals is determined. The data set is analyzed to find the minimum canopy openness that can achieve stable satellite signal reception and successful positioning. This value is set as the minimum threshold of canopy openness.

[0153] The candidate points under the forest where the canopy openness is greater than the minimum threshold are located, and the drone hovering flight positioning experiments with durations of 1min, 30s, 20s, 10s, 5s, and 3s are conducted;

[0154] Furthermore, the positioning of candidate points under the forest where the canopy openness is greater than the minimum threshold is performed by conducting a drone hovering flight positioning experiment with a duration of 1 minute, 30 seconds, 20 seconds, 10 seconds, 5 seconds, and 3 seconds, including:

[0155] At the candidate points, the drone is controlled to hover for the designed duration and positioned through the received satellite signals; multiple candidate points are selected for experiments in areas where the canopy openness is greater than the previously determined minimum threshold. Ensure that each point meets the basic requirements for satellite signal reception. Six different hovering durations are set, namely 1 minute, 30 seconds, 20 seconds, 10 seconds, 5 seconds, and 3 seconds. These durations will be used to evaluate the impact of hovering time on positioning performance.

[0156] In each experiment, record whether the drone can successfully obtain a fixed positioning solution and the specific time required for positioning; at each candidate point, control the drone to hover for the designed duration. During the hovering period, the drone is positioned through the received satellite signals.

[0157] Analyze the impact of different hovering times on the positioning success rate, positioning speed and positioning accuracy to evaluate the positioning performance of the UAV under different hovering times;

[0158] Get the shortest hovering time required by the drone while ensuring positioning accuracy.

[0159] Record and analyze the situation of the drone achieving a fixed position solution in each flight time period under different canopy openness; in each experiment, record whether the drone can successfully obtain a fixed position solution. Record the specific time required for positioning, that is, the time interval from the drone starting to hover to successfully obtaining a fixed position solution.

[0160] Based on the experimental data, the impact of different hovering times on positioning accuracy is evaluated, and the optimal solution for the hovering time and positioning accuracy of the drone is obtained;

[0161] According to the best solution, satellite signals are used to hover and locate the base station drone under the forest.

[0162] The fixed positioning solution of the UAV is a high-precision positioning state achieved during the UAV RTK (Real Time Kinetic) positioning process. RTK technology combines the global navigation satellite system (such as GPS, GLONASS, Galileo, BDS, etc.) with real-time differential technology, and achieves centimeter-level or even millimeter-level positioning accuracy through real-time data transmission and processing between the ground base station and the rover on the UAV.

[0163] In actual applications, when positioning a drone, a longer hovering time may improve positioning accuracy, but it will also increase energy consumption and time costs. Therefore, it is necessary to find a balance between positioning accuracy and hovering time. The best solution for this application should take into account the following factors:

[0164] (1) Positioning accuracy: Ensure that the UAV can achieve the required positioning accuracy during hovering.

[0165] (2) Hovering time: Shorten the hovering time as much as possible to reduce energy consumption and time cost.

[0166] (3) Signal stability: Ensure that the drone can stably receive satellite signals during hovering.

[0167] S300, positioning the understory sensing drone cluster based on the location information of the base station drone;

[0168] Furthermore, the positioning of the understory sensing drone cluster based on the location information of the base station drone includes:

[0169] By measuring the distance or angle data between the base station drone and the perception drone, the position of the perception drone is determined using a geometric algorithm;

[0170] Using the communication and measurement equipment between the base station drone and the perception drone, the distance or angle data between the base station drone and each perception drone is collected in real time;

[0171] Use the distance resection method or free station method to calculate the flight trajectory and state parameters of each sensing drone in real time based on the collected distance or angle data.

[0172] In this step, the base station drone is used to sense and locate the understory sensing drone cluster. First, the distance or angle measurement data between the base station drone and the sensing drone needs to be obtained. These data can be obtained through various sensors on the drone, such as GNSS receivers, radars, etc. Then, based on the acquired data, geometric positioning algorithms such as distance resection or free station setting are used to solve the flight trajectory and state parameters of the sensing drone in real time. These algorithms can process data in real time to ensure the rapid update of the position of the sensing drone, and at the same time combine the position information of multiple base station drones to improve the positioning accuracy. In this way, each member of the sensing drone cluster can use the position information of the base station drone to locate itself, thus forming a collaborative network. This not only provides a positioning benchmark for the sensing drone, but also provides important support for its mission planning, path optimization, etc. This method effectively solves the problem of weak or no GPS signal in the understory environment, improves the operating efficiency and safety of drones under the forest, and can achieve wider area coverage and more efficient data collection through the collaborative work of multiple drones.

[0173] S400, based on the collaboration of multiple drones of base station drones and perception drone clusters, explores the positioning signals under the forest and realizes the collaborative positioning of multiple drones under the forest.

[0174] Furthermore, the multi-UAV collaboration based on the base station UAV and the perception UAV cluster explores the understory positioning signal and realizes the multi-UAV collaborative understory positioning, including:

[0175] In the process of multi-UAV coordinated forest positioning, a dynamic rotation mechanism of the "leader" base station UAV group is established to dynamically alternate the identities of the base station UAV and the perception UAV;

[0176] Among them, after the perception drone finds the candidate positioning point area under Helin, it can be converted into a base station drone, and perform rapid static measurement and positioning with other base station drones, autonomously exploring and collaborating with each other;

[0177] The base station drone adjusts its altitude up and down when hovering to obtain the best satellite positioning signal, and the sensing drone flies at a fixed reference altitude.

[0178] Furthermore, the construction of a dynamic rotation mechanism for the “leading goose” base station drone group, which dynamically alternates the identities of the base station drone and the perception drone, includes:

[0179] In the process of multiple drones coordinating under-forest positioning, the “leader” base station drone group uses edge connection to form a network, forming a strip network composed of triangles or geodetic quadrilaterals;

[0180] During the dynamic rotation of the base station drone group, at least three base station drones are ensured to be in a hovering positioning state during the station relocation process;

[0181] Each time the station is moved, the base station drone moves forward to explore satellite positioning signals based on its own navigation trajectory and real-time positioning results.

[0182] In the scheme claimed for protection in this application, in the process of collaborative positioning of multiple drones, a "leader" drone group consisting of several base station drones is constructed, and a network is constructed in an edge connection manner to ensure communication and data exchange. The dynamic rotation mechanism ensures that at least three base station drones are always in a hovering positioning state to ensure the continuity and stability of positioning. After finding a suitable point, the perception drone can be converted into a base station drone and perform rapid static measurement and positioning with other base station drones. This alternation of identities improves the flexibility and positioning accuracy of the system. The base station drone dynamically adjusts its flight altitude according to the quality of the received satellite signal to capture the optimal satellite signal and improve positioning accuracy. The perception drone maintains a fixed altitude flight to ensure stable and continuous perception. Through the dynamic rotation mechanism, the system always has a sufficient number of base station drones for positioning, greatly improving positioning efficiency.

[0183] The application claims a method for collaborative positioning of multiple UAVs in an understory economic zone. Through the hovering flight mode of multiple UAVs, the hovering time is adjusted in real time according to different canopy openness conditions, thereby effectively improving the positioning efficiency in a complex understory environment. In addition, this method can use GNSS satellite signals such as Beidou to assist in positioning, and combine algorithms to quickly calculate the canopy openness, optimize the hovering time and positioning strategy of the UAV, which can improve the final positioning accuracy; in addition, this method uses the collaborative work of multiple UAVs of base station UAVs and perception UAV clusters for positioning, which improves the fault tolerance and self-healing ability of the overall system. In summary, this method can achieve efficient and accurate positioning of the understory economic zone, promote integrated location information services above and below the canopy, provide mechanization and automation support for understory planting and breeding, and help the intensive and efficient development of the understory economy. This method has a wide range of applications. In addition to being applicable to understory economic zones, it can also be widely used in UAV collaborative positioning tasks in other complex environments, such as mountainous areas, forests, urban canyons, etc.

[0184] For ease of description, spatially relative terms such as "above", "above", "on the upper surface of", "above", etc. may be used here to describe the spatial positional relationship between a device or feature and other devices or features as shown in the figure. It should be understood that spatially relative terms are intended to include different orientations of the device in use or operation in addition to the orientation described in the figure. For example, if the device in the accompanying drawings is inverted, the device described as "above other devices or structures" or "above other devices or structures" will be positioned as "below other devices or structures" or "below other devices or structures". Thus, the exemplary term "above" can include both "above" and "below". The device can also be positioned in other different ways (rotated 90 degrees or in other orientations), and the spatially relative descriptions used here are interpreted accordingly.

[0185] In addition, it should be noted that the use of terms such as "first" and "second" to limit components is only for the convenience of distinguishing the corresponding components. Unless otherwise stated, the above terms have no special meaning and therefore cannot be understood as limiting the scope of protection of this application.

[0186] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for collaborative positioning of multiple UAVs in forest economic zones, characterized in that: include: Obtain full-sky photos of forest understory points and calculate the canopy openness of the points; Set the drone hovering positioning time according to the canopy width, and use satellite signals to perform base station drone hovering positioning under the forest; Based on the location information of the base station drone, the understory sensing drone cluster is positioned; Based on the collaboration of multiple UAVs including base station UAVs and perception UAV clusters, we explore the understory positioning signals and realize the collaborative understory positioning of multiple UAVs.

2. The method for coordinating multiple UAVs in forest economic zones according to claim 1 is characterized by: The method of obtaining a full sky photo of a point under the forest and calculating the canopy openness of the point includes: Using a photosensitive element to trigger a camera module to take a photo, obtaining a full-sky photo of a forest understory point that can receive a satellite signal; wherein the satellite signal is a Beidou satellite signal; Calculate the canopy openness of a point based on a deep learning algorithm.

3. The method for coordinating multiple UAVs in forest economic zones according to claim 2 is characterized in that: The method of using a photosensitive element to trigger a camera module to take photos, and obtaining full-sky photos of forest locations that can receive satellite signals, includes: The threshold of sunlight brightness change per unit time is selected as the criterion for selecting candidate locations for understory positioning; Use photosensitive elements to monitor the light intensity on the top of the drone in real time. When the photosensitive element detects that the light intensity exceeds the set threshold, the circular camera module is triggered to start taking pictures. During the photo shooting process, use a wide-angle fisheye lens to shoot photos with a hemispherical field of view greater than 150°; Correct the captured fisheye image to eliminate the distortion caused by the fisheye lens. Record and store corrected full-sky images.

4. The method for coordinating multiple UAVs in forest economic zones according to claim 3 is characterized in that: The threshold value of sunlight brightness change per unit time is selected as the standard for selecting candidate locations for understory positioning, including: The sunlight brightness is calculated using the reverse ray tracing method; specifically, the light starts from the camera, and the light entering the camera is traced back to calculate the brightness of each point in the camera's field of view; by assuming that the surface of the object is an ideal mirror reflection, the reflected light is traced to the light source, and the scattering distribution during the light transmission process is recorded, thereby deriving the basic formula for the brightness calculation equation: Among them, ω i is the spatial solid angle of the i-th ray, φ is the angular range of the spatial solid angle, L ei (P,ω0,λ) is the incident spectral radiance, θ i is the angle between the incident light ray and the normal element; f γ (P,ω i ,ω0,λ) is the spectral reflectivity of the object surface to incident light in any direction in any direction. The specific expression is as follows: Using the basic formula of the brightness calculation equation, calculate the spectral radiance data of each point on the camera's viewing plane, and calculate the camera's radiance value and the brightness value perceived by the camera; The camera's radiation brightness value is calculated as follows: The brightness value perceived by the camera is calculated as follows: Among them, K m It is expressed as the maximum spectral visual efficacy under bright vision conditions, and V(λ) represents how many lumens of single-spectrum luminous flux can be converted from single-spectrum radiant flux per watt; The brightness information in the forest environment is obtained according to the camera's radiation brightness value and the brightness value perceived by the camera, so as to calculate the sunlight brightness per unit time and the threshold value of sunlight brightness change per unit time; When the light intensity at the top of the drone increases above this threshold, the point is considered a candidate point for understory positioning.

5. The method for coordinating multiple UAVs in forest economic zones according to claim 2 is characterized in that: The canopy openness of the point calculated based on the deep learning algorithm includes: The canopy openness refers to the proportion of the sky sphere that is not blocked by tree branches when looking up from a point in the forest. The calculation of canopy openness is achieved by calculating canopy closure, where canopy closure is the ratio of the projected area of ​​the canopy to the forest area; The relationship between canopy openness and canopy closure is as follows: Canopy openness = 1-canopy closure; The calculation process of canopy closure is as follows: Obtain the full sky fisheye image of the forest understory point, and set the actual area represented by each pixel under different resolution conditions to f(d), that is: Pa=f(d) Where Pa is the actual area represented by the pixel in the image; d is the resolution of the scanner; Calculate the positive area S of the leaf in the acquired image. The specific calculation formula is as follows: Among them, ∮ S d(P) is the integral of the number of pixels contained in the leaf image contour in the acquired image; The canopy closure is calculated using the leaf orthophoto area S and the total area of ​​the acquired image; or, The calculation process of canopy closure is as follows: Obtain all-sky fisheye images of understory locations; digitizing the acquired image, and calculating the number of pixels contained in the plant leaf image from the digitized image; The canopy closure is calculated using the number of pixels contained in the plant leaf image and the number of pixels contained in the entire image.

6. The method for coordinating multiple UAVs in forest economic zones according to claim 1 is characterized by: The method of setting the drone hovering positioning time according to the canopy width and using satellite signals to perform under-forest hovering positioning of the base station drone includes: Using the intelligent perception of photosensitive elements, multiple candidate locations for understory positioning were selected, and a 5-minute drone hovering flight positioning experiment was conducted at the candidate locations. The canopy openness of the points where the drone can hover and locate within 5 minutes is counted, and based on the statistics, the minimum threshold of canopy openness for drone understory positioning that can obtain satellite signal support is determined; The candidate points under the forest where the canopy openness is greater than the minimum threshold are located, and the drone hovering flight positioning experiments with durations of 1min, 30s, 20s, 10s, 5s, and 3s are conducted; Record and analyze the situation of the drone achieving a fixed positioning solution in each flight time period under different canopy openness; Based on the experimental data, the impact of different hovering times on positioning accuracy is evaluated, and the optimal solution for the hovering time and positioning accuracy of the drone is obtained; According to the best solution, satellite signals are used to hover and locate the base station drone under the forest.

7. The method for coordinating multiple UAVs in forest economic zones according to claim 6 is characterized by: The candidate points under the forest where the canopy openness is greater than the minimum threshold are located, and the drone hovering flight positioning experiment with a duration of 1 minute, 30 seconds, 20 seconds, 10 seconds, 5 seconds, and 3 seconds is conducted, including: Control the drone to hover at the candidate point for a designed duration and locate it through the received satellite signals; In each experiment, record whether the drone can successfully obtain a fixed positioning solution and the specific time required for positioning; Analyze the impact of different hovering times on the positioning success rate, positioning speed and positioning accuracy to evaluate the positioning performance of the UAV under different hovering times; Get the shortest hovering time required by the drone while ensuring positioning accuracy.

8. The method for collaborative positioning of multiple UAVs in forest economic zones according to any one of claims 1 to 7, characterized in that: The positioning of the understory sensing drone cluster based on the location information of the base station drone includes: By measuring the distance or angle data between the base station drone and the perception drone, the position of the perception drone is determined using a geometric algorithm; Using the communication and measurement equipment between the base station drone and the perception drone, the distance or angle data between the base station drone and each perception drone is collected in real time; Use the distance resection method or free station method to calculate the flight trajectory and state parameters of each sensing drone in real time based on the collected distance or angle data.

9. The method for coordinating multiple UAVs in forest economic zones according to claim 8 is characterized in that: The multi-UAV collaboration based on the base station UAV and the perception UAV cluster explores the understory positioning signal and realizes the multi-UAV collaborative understory positioning, including: In the process of multi-UAV coordinated forest positioning, a dynamic rotation mechanism for the base station UAV group is constructed to dynamically alternate the identities of the base station UAV and the perception UAV; Among them, after the perception drone finds the candidate positioning point area under Helin, it can be converted into a base station drone, and perform rapid static measurement and positioning with other base station drones, autonomously exploring and collaborating with each other; The base station drone adjusts its altitude up and down when hovering to obtain the best satellite positioning signal, and the sensing drone flies at a fixed reference altitude.

10. The method for coordinating multiple UAVs in forest economic zones according to claim 9 is characterized in that: The construction of a dynamic rotation mechanism for the base station drone group, dynamically alternating the identities of the base station drone and the perception drone, includes: In the process of multi-UAV coordinated forest positioning, the base station UAV group uses edge connection to form a network, forming a strip network composed of triangles or geodetic quadrilaterals; During the dynamic rotation of the base station drone group, at least three base station drones are ensured to be in a hovering positioning state during the station relocation process; Each time the station is moved, the base station drone moves forward to explore satellite positioning signals based on its own navigation trajectory and real-time positioning results.

Images