An automatic blueberry picking robot

Abstract

This invention relates to an automated blueberry harvesting robot, belonging to the field of harvesting machinery technology. It includes: a moving mechanism, a controller, a blueberry recognition mechanism, a multi-axis robotic arm mechanism, and a blueberry harvesting mechanism. The controller is fixed to the moving mechanism. The blueberry recognition mechanism includes a LiDAR and a color camera; the LiDAR is fixed to the controller; the color camera is fixed to the LiDAR. The multi-axis robotic arm mechanism is fixed to the moving mechanism. The blueberry harvesting mechanism is detachably connected to the multi-axis robotic arm mechanism. This invention utilizes the controller in conjunction with the blueberry recognition mechanism to collect data, build a map, and plan a path, enabling autonomous navigation. The controller then fuses the point cloud data collected by the LiDAR and the color images captured by the color camera to colorize the point cloud, which is then fed into a pre-trained fcaf3d network to accurately locate ripe blueberries and determine their growth direction. Finally, the multi-axis robotic arm mechanism drives the blueberry harvesting mechanism to move along a set route for rapid blueberry harvesting.

Description

Technical Field

[0001] This invention relates to the field of harvesting machinery technology, and in particular to an automatic blueberry harvesting robot. Background Technology

[0002] Blueberries are charming small berries with a beautiful blue color and an intoxicating aroma. Their flesh is delicate, with a perfectly balanced sweet and tart flavor, making them a mouthwatering delicacy. The edible portion of the blueberry makes up the entire fruit, which is not only delicious but also rich in nutrients, earning it the reputation of a natural treasure trove. Its unique antioxidant properties make it an ideal choice for preventing brain aging, maintaining eyesight, enhancing heart health, fighting cancer, promoting vascular health, and boosting the body's immunity.

[0003] Blueberries are rich in a variety of nutrients, including antioxidants such as vitamin C, vitamin E, and flavonoids. These antioxidants help neutralize free radicals, reduce oxidative stress, and protect cells from damage. Furthermore, the anthocyanins in blueberries are powerful antioxidants with anti-inflammatory and anti-cancer effects. They also help improve memory and learning ability, and protect brain health. Blueberries are also rich in dietary fiber, which helps promote digestive health, prevent constipation, and help control blood sugar and cholesterol levels. In addition, blueberries contain various vitamins and minerals, such as vitamin K, vitamin C, manganese, and iron, which are essential for maintaining the immune system, bone health, and red blood cell production.

[0004] Blueberries, a delicious, healthy, and valuable economic crop, are widely welcomed in the international market. However, the labor-intensive and technically demanding nature of blueberry harvesting results in high costs and limited efficiency. Currently, blueberry harvesting relies primarily on manual labor, which not only leads to cost issues but also restricts the sustainable development of the blueberry industry.

[0005] Existing technologies have attempted to improve blueberry harvesting efficiency using simple auxiliary tools, such as forked steel forks. However, these tools still present several problems. First, because blueberries are small and fragile, using rough tools can damage and cause the fruit to fall off. Second, these simple tools are often unsuitable for the different growth patterns and fruit distributions of blueberry bushes, limiting their applicability in various environments.

[0006] Furthermore, the short ripening period of blueberries limits the harvesting time, requiring a large volume of harvesting to be completed within a limited timeframe. The limitations of manual harvesting hinder the full increase in blueberry yield, and the shortage of human resources exacerbates this problem, restricting the expansion of the blueberry industry.

[0007] Therefore, there is a need in the existing technology for an innovative blueberry harvesting device to overcome the aforementioned problems, improve harvesting efficiency, reduce harvesting losses, and enable mechanized harvesting in the blueberry industry. Such a device needs to be able to accurately identify blueberry fruits, gently harvest them without causing damage, adapt to different blueberry bush shapes and fruit distributions, and meet the seasonal demand for efficient harvesting. This device can reduce blueberry harvesting costs and enhance the competitiveness and sustainability of the blueberry industry. Summary of the Invention

[0008] This invention provides an automated blueberry harvesting robot, which solves the technical problems of low blueberry harvesting efficiency and high blueberry damage rate.

[0009] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: an automatic blueberry harvesting robot, comprising: a moving mechanism, a controller, a blueberry recognition mechanism, a multi-axis robotic arm mechanism, and a blueberry picking mechanism.

[0010] The controller is fixed to the moving mechanism; the blueberry recognition mechanism includes a lidar and a color camera, the lidar being fixed to the controller; the color camera being fixed to the lidar; the multi-axis robotic arm is located on one side of the controller and fixed to the moving mechanism; the blueberry picking mechanism is detachably connected to the multi-axis robotic arm; the controller is electrically connected to the lidar, the color camera, the multi-axis robotic arm, and the blueberry picking mechanism respectively.

[0011] The beneficial effects of this invention are:

[0012] 1. By using a controller combined with a blueberry recognition device (LiDAR) to collect data, a map is created, and a route is planned on the created map, autonomous navigation can be achieved;

[0013] 2. The controller fuses the point cloud data collected by the LiDAR and the color images captured by the color camera to colorize the point cloud. Then, it is fed into a pre-trained 3D point cloud recognition network (FCAF3D network) to accurately obtain the spatial location of ripe blueberries.

[0014] 3. By using a controller to autonomously adjust the multi-axis robotic arm mechanism to move the blueberry picking mechanism in space, blueberries can be collected autonomously, improving blueberry harvesting efficiency.

[0015] 4. The blueberry picking mechanism can be detachably connected to the multi-axis robotic arm mechanism, and different widths of blueberry picking mechanisms can be selected according to different blueberry tree species, effectively improving the applicability of the blueberry picking robot.

[0016] Based on the above technical solution, the present invention can be further improved as follows.

[0017] Furthermore, the moving mechanism includes a carriage, a moving motor, drive wheels, driven wheels, tracks, and a battery. There are two moving motors, both fixed inside the carriage, with their output shafts extending out from opposite sides of the carriage. There are two drive wheels located outside the carriage and respectively mounted on the output shafts of the two moving motors. Multiple driven wheels are rotatably connected to opposite outer sides of the carriage. Two tracks are correspondingly and tactilely connected to the drive wheels and multiple driven wheels on both sides of the carriage. The battery is placed inside the carriage. The controller is fixed to the carriage and electrically connected to the moving motors and the battery. The multi-axis robotic arm mechanism is fixed to the carriage.

[0018] The further beneficial effects of adopting the above are: tracks are more suitable for complex environments such as unevenness, mud, and ruggedness in orchards than wheeled vehicles. They can provide a larger ground contact area, reduce pressure on the ground, thereby reducing the risk of traps on soft soil and other loose ground. They can also provide greater traction, higher stability, and greater load capacity.

[0019] Furthermore, the controller includes a control cabinet, an inertial sensor, and a main control computer embedded with the LIO-SAM mapping algorithm. The bottom of the control cabinet is fixed to the top of the vehicle body; the inertial sensor is fixed inside the control cabinet; the main control computer is fixed inside the control cabinet and electrically connected to the moving motor, the inertial sensor, the lidar, the color camera, the multi-axis robotic arm mechanism, and the blueberry picking mechanism; the lidar is fixed to the vehicle body.

[0020] The further beneficial effects of adopting the above are as follows: The main control computer fuses the data transmitted from the LiDAR and inertial sensors to build a map, uses the LIO-SAM mapping algorithm to create a high-definition point cloud map, performs a series of filtering and segmentation operations on the built map to remove obstacles and retain only passable areas, and then uses a path planning algorithm to plan the forward path to achieve autonomous navigation; at the same time, the point cloud data collected by the LiDAR and the color images taken by the color camera are fused to colorize the point cloud, and then fed into a pre-trained 3D point cloud recognition network (FCAF3D network) to determine the location of the blueberries, obtain the blueberry production direction, and then, through the calculation of the main control computer, control the moving mechanism to move to the mature blueberry growing location according to the set path.

[0021] Furthermore, the blueberry picking mechanism includes a fixed box, a picking motor, a drive wheel, driven wheels, a wheel rim, a fixed rod, a moving rod, and a picking box. The fixed box is detachable from the multi-axis robotic arm mechanism. The interior of the fixed box has a cavity arranged parallel to the height direction, and the longitudinal section of the cavity is circular. The side of the fixed box away from the vehicle body has a movable outlet hole communicating with the cavity. The picking motor is fixed to the outer side of the fixed box, and its output shaft extends into the fixed box. The output shaft of the picking motor is parallel to the axis of the cavity. The drive wheel is located inside the cavity and is sleeved on the output shaft of the picking motor. Multiple driven wheels are arranged in a planetary pattern on the drive wheel. On the outer periphery, multiple driven wheels are rotatably connected to the cavity wall of the cavity and are all meshed with the driving wheel; the wheel rim is rotatably connected to the cavity along the circumference of the cavity, and the inner surface of the wheel rim is fixed with gear teeth, which are all meshed with the multiple driven wheels; the fixed rod is arranged parallel to the fixed box, and one end of the rod is fixed to the outer side of the fixed box corresponding to the movable outlet hole, and the other end of the rod is fixed with a sliding sleeve; the moving rod is arranged along the length of the fixed rod, and one end of the rod is fixed to the wheel rim, and the other end of the rod slides through the movable outlet hole and the sliding sleeve; the picking box is fixed to the other end of the moving rod, and the other end of the moving rod away from the vehicle box is provided with a picking hole.

[0022] The further beneficial effects of adopting the above are:

[0023] 1. An innovative improvement to the traditional crank-rocker mechanism, replacing the crank with a planetary gear train, can increase the motor transmission ratio, increase the transmission torque, and improve the blueberry harvesting efficiency.

[0024] 2. By using the rotating wheel to drive the moving rod to slide back and forth along the sliding sleeve, the picking box can move regularly along the elliptical axis, realizing hand-like picking, quickly removing blueberries, improving picking efficiency and reducing blueberry damage rate.

[0025] Furthermore, the picking box includes a baffle, a bottom plate, a top plate, comb teeth, and side plates. The baffle is vertically fixed to the other end of the moving rod. The bottom plate and the top plate are fixed vertically to the side of the baffle away from the vehicle body. There are multiple comb teeth, all arranged along the length of the moving rod, with one end of each comb tooth fixed at a distance from the end of the bottom plate away from the baffle. There are two side plates, both arranged along the length of the moving rod, with the two side plates fixed to the sides of the bottom plate and the top plate, and each side plate is fixed to an adjacent comb tooth. The picking hole is the gap between the top plate, the multiple comb teeth, and the two side plates.

[0026] The further beneficial effect of adopting the above method is that using the comb teeth to scrape off ripe blueberries can improve the efficiency of blueberry collection and reduce the blueberry harvesting damage rate.

[0027] Furthermore, the picking box also includes a blueberry transport tube, and the inner bottom surface of the base plate is provided with a downward slope from the edge to the center and a blueberry transport hole at its lowest point; one end of the blueberry transport tube is connected to the blueberry transport hole and the other end extends into the box.

[0028] The further beneficial effect of adopting the above is that: after the blueberries are picked, they are transported to the truck bed through the blueberry transport pipe for storage, which can increase the harvesting time of the blueberry harvesting robot.

[0029] Furthermore, all of the comb teeth are made of rubber.

[0030] The further beneficial effect of adopting the above is that using rubber material to make the comb teeth can avoid damaging the blueberries and reduce the damage rate during blueberry harvesting.

[0031] Furthermore, the diameter of each of the comb teeth decreases sequentially in the direction away from the base plate.

[0032] The further beneficial effect of adopting the above is that the action of multiple comb teeth decreases sequentially in the direction away from the base plate, which can not only improve the picking efficiency of comb tooth harvesting, but also prevent blueberries from falling off the comb teeth. Attached Figure Description

[0033] Figure 1 This is a three-dimensional structural diagram of an automatic blueberry harvesting robot according to the present invention;

[0034] Figure 2 This is a front view structural diagram of an automatic blueberry harvesting robot according to the present invention;

[0035] Figure 3 This is a front view schematic diagram of the blueberry picking mechanism in an automatic blueberry harvesting robot according to the present invention;

[0036] Figure 4 This is a rear view schematic diagram of the blueberry picking mechanism in an automatic blueberry harvesting robot according to the present invention;

[0037] Figure 5 This is a schematic diagram of an automatic blueberry harvesting robot according to the present invention.

[0038] The attached diagram lists the components represented by each number as follows:

[0039] 1. Moving mechanism; 11. Carriage box; 12. Track; 2. Controller; 21. Control cabinet; 22. Inertial sensor; 23. Main control computer; 3. Blueberry recognition mechanism; 31. LiDAR; 32. Color camera; 4. Multi-axis robotic arm mechanism; 5. Blueberry picking mechanism; 51. Fixed box; 52. Picking motor; 53. Drive wheel; 54. Driven wheel; 55. Wheel rim; 56. Fixed rod; 57. Moving rod; 58. Picking box; 581. Top plate; 582. Comb teeth; 583. Side plate; 59. Blueberry transport pipe. Detailed Implementation

[0040] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0041] like Figure 1 As shown, an automated blueberry harvesting robot includes: a mobile mechanism 1, a controller 2, a blueberry recognition mechanism 3, a multi-axis robotic arm mechanism 4, and a blueberry picking mechanism 5.

[0042] The controller 2 is fixed to the moving mechanism 1; the blueberry recognition mechanism 3 includes a lidar 31 and a color camera 32, with the lidar 31 fixed to the controller 2 and the color camera 32 fixed to the lidar 31; the multi-axis robotic arm mechanism 4 is located on one side of the controller 2 and fixed to the moving mechanism 1; the blueberry picking mechanism 5 is detachably connected to the end effector of the multi-axis robotic arm mechanism 4; the controller 2 is electrically connected to the lidar 31, the color camera 32, the multi-axis robotic arm mechanism 4, and the blueberry picking mechanism 5 respectively.

[0043] In some specific embodiments, the moving mechanism 1 may include a vehicle body 11, a moving motor, a driving pulley, a driven pulley, a track 12, and a battery. There are two moving motors, both fixed inside the vehicle body 11, with the output shafts of the two moving motors extending out from opposite sides of the vehicle body 11. There are two driving pulleys located outside the vehicle body 11 and respectively mounted on the output shafts of the two moving motors. There are multiple driven pulleys that are rotatably connected to opposite outer sides of the vehicle body 11 at intervals. There are two tracks 12 that are correspondingly and tactilely connected to the driving pulleys and multiple driven pulleys on both sides of the vehicle body 11. The battery is placed inside the vehicle body 11. The controller 2 is fixed on the vehicle body 11 and electrically connected to the moving motors and the battery. The multi-axis robotic arm mechanism 4 is fixed on the vehicle body 11.

[0044] In some specific embodiments, the controller 2 may include a control cabinet 21, an inertial sensor 22, and a main control computer 23 with an embedded LIO-SAM mapping algorithm. The bottom surface of the control cabinet 21 is fixed to the top surface of the vehicle body 11; the inertial sensor 22 is fixed inside the control cabinet 21; the main control computer 23 is fixed inside the control cabinet 21 and is electrically connected to the moving motor, the inertial sensor 22, the lidar 31, the color camera 32, the multi-axis robotic arm mechanism 4, and the blueberry picking mechanism 5 respectively; the lidar 31 is fixed on the vehicle body 11.

[0045] In some specific embodiments, the blueberry picking mechanism 5 may include a fixed box 51, a picking motor 52, a drive wheel 53, a driven wheel 54, a wheel rim 55, a fixed rod 56, a moving rod 57, and a picking box 58. The fixed box 51 is detachably connected to the end effector of the multi-axis robotic arm mechanism 4. The interior of the fixed box 51 has a cavity arranged parallel to the height direction, and the longitudinal section of the cavity is circular. The side of the fixed box 51 away from the carriage 11 has a movable outlet hole communicating with the cavity. The picking motor 52 is fixed to the outer side of the fixed box 51, and its output shaft extends into the fixed box 51. The output shaft of the picking motor 52 is parallel to the axis of the cavity. The drive wheel 53 is located in the cavity and is sleeved on the output shaft of the picking motor 52. The driven wheel 54... Multiple driven wheels 54 are arranged in a planetary pattern on the outer periphery of the driving wheel 53. Multiple driven wheels 54 are rotatably connected to the cavity wall and mesh with the driving wheel 53. A wheel rim 55 is rotatably connected to the cavity along the circumference of the cavity. The inner surface of the wheel rim 55 is fixed with teeth, and the teeth mesh with the multiple driven wheels 54. A fixed rod 56 is arranged parallel to the fixed box 51, and one end of the rod is fixed to the outer side of the fixed box 51 corresponding to the movable outlet hole. The other end of the rod is fixed with a sliding sleeve. A moving rod 57 is arranged along the length of the fixed rod 56. One end of the moving rod 57 is fixed to the wheel rim 55, and the other end slides through the movable outlet hole and the sliding sleeve. A picking box 58 is fixed to the other end of the moving rod 57, and the other end of the moving rod 58 away from the box 11 is provided with a picking hole.

[0046] In some specific embodiments, the picking box 58 may include a baffle, a bottom plate, a top plate 581, comb teeth 582, and side plates 583. The baffle is vertically fixed to the other end of the moving rod 57. The bottom plate and the top plate 581 are fixed vertically to the side of the baffle away from the carriage 11. There are multiple comb teeth 582, all arranged along the length of the moving rod 57. One end of the multiple comb teeth 582 is fixed at a distance from the end of the bottom plate away from the baffle. There are two side plates 583, both arranged along the length of the moving rod 57. The two side plates 583 are fixed to the two sides of the bottom plate and the top plate 581, and both side plates 583 are fixed to the adjacent comb teeth 582. The picking hole is the gap between the top plate 581, the multiple comb teeth 582, and the two side plates 583.

[0047] In some specific embodiments, the picking box 58 may also include a blueberry transport tube 59, the inner bottom surface of the base plate is provided with a downward slope from the edge to the center and a blueberry transport hole at its lowest point; one end of the blueberry transport tube 59 is connected to the blueberry transport hole and the other end is inserted into the carriage 11.

[0048] In some specific embodiments, multiple comb teeth 582 can be made of rubber material.

[0049] In some specific embodiments, the diameter of the plurality of comb teeth 582 decreases sequentially in the direction away from the base plate.

[0050] Example

[0051] The automated blueberry harvesting robot, controlled by a remote, moves throughout the blueberry orchard, traversing all the paths before returning to its starting point. Throughout this process, the LiDAR 31, inertial sensor 22, and main control computer 23 play crucial roles. During the robot's movement, the LiDAR 31 continuously performs a 360-degree panoramic scan of the surrounding orchard environment. Simultaneously, data from the inertial sensor 22 is transmitted to the main control computer 23, where the LIO-SAM algorithm is activated. In the current mapping phase, the LIO-SAM algorithm performs point cloud distortion correction and feature extraction. Point cloud distortion correction corrects the LiDAR 31 data by establishing an extrinsic parameter model and a motion distortion model, thus obtaining accurate point cloud information. Feature extraction utilizes the geometric and intensity information of feature points to reduce computational and storage requirements. Simultaneously, LIO-SAM utilizes the attitude information provided by inertial sensor 22 for motion prediction. By integrating the data from inertial sensor 22, the robot's pose changes can be estimated and used as an initial estimate for the scan matching process. Next, the algorithm uses a scan matching method, comparing the feature point positions of the current frame with those of the previous frame to infer the robot's pose changes. Then, the LIO-SAM algorithm maps these feature points onto a map, progressively constructing an environmental map. This feature-based matching method effectively solves the data association problem and improves the accuracy and robustness of the mapping.

[0052] The robot is remotely controlled to navigate the entire blueberry orchard and then returns to its starting position, completing a closed loop. Upon detecting this loop, the LIO-SAM algorithm optimizes it by adjusting the robot's position and posture on the map to ensure a more consistent map topology at the loop point, thus improving the overall map accuracy and consistency. After building the map, the algorithm further identifies roads and obstacles to prepare for the next step of path planning.

[0053] After the mapping process for the automated blueberry harvesting robot, the next step is to activate its automatic control function. Once activated, the robot will autonomously plan its path based on the obstacles and passable roads defined in the previously created high-precision map.

[0054] As the robot moves along its pre-planned path, the LiDAR 31 and inertial sensor 22 remain on, and the color camera 32 is also activated. The LiDAR 31 continuously scans the surrounding environment, sending point cloud data of the environment to the main control computer 23, while the inertial sensor 22 continuously transmits its measured attitude information.

[0055] At this time, the LIO-SAM algorithm is also running, in the localization phase. LIO-SAM uses the established map and the current scan data from LiDAR 31 to estimate the robot's pose through least-squares optimization. During this process, LIO-SAM uses the attitude information provided by inertial sensor 22 for motion compensation, correcting motion distortions in the LiDAR 31 data. By fusing the data from inertial sensor 22 and LiDAR 31, LIO-SAM can more accurately estimate the robot's attitude changes and trajectory during movement. Finally, LIO-SAM outputs the robot's pose and map information.

[0056] During the autonomous navigation process of the blueberry automatic harvesting robot, the sent point cloud data is segmented. Specifically, the point cloud data of the left side of the moving mechanism 1, i.e., the location of the blueberry picking mechanism 5 at the end, and the ring point cloud data are segmented. The segmented point cloud data is then fused with the color image captured by the color camera to transform the ordinary point cloud into a color point cloud, which can effectively improve the recognition rate of ripe blueberries. The color point cloud is then transmitted into the trained 3D point cloud detection network (FCAF3D).

[0057] Before being fed into the 3D point cloud detection network, the colored point cloud undergoes a preprocessing stage, including data calibration, noise reduction, and spatial transformation to adapt to the network input format. Subsequently, the feature extraction module in the network uses an advanced deep learning architecture to capture local geometric features and global contextual information of the point cloud at multiple scales. Specifically, FCAF3D enables each point to dynamically weight and fuse its neighborhood features based on global information, significantly improving the representation ability of 3D objects in complex scenes. After feature fusion, the network enters the prediction stage, no longer relying on traditional anchor box design, but directly predicting the class probability, bounding box regression parameters, and centrality for each location. This covers the center point's position offset, object size, and pose estimation. Finally, non-maximum suppression (NMS) is applied to eliminate redundant detection results, thus determining the location of the ripe blueberry.

[0058] Through prediction and reasoning using a 3D point cloud detection network, the position of a ripe blueberry relative to a blueberry harvesting robot can be obtained.

[0059] The process of a blueberry harvesting robot identifying ripe blueberries is as follows: Figure 5 As shown in the diagram. First, the lidar 31 continuously scans the surrounding environment to acquire point cloud data, while the color camera 32 continuously captures images of the opposite environment to acquire color image data. The data flow is as follows: Figure 5 As indicated by the arrow below, after acquiring point cloud data and color image data of the surrounding objects, the data is transmitted to the main control computer 23 below the lidar 31. After color fusion, the data is input into the 3D point cloud detection network. The location of the ripe blueberry is then determined through inference. The main control computer 23 then performs calculations and coordinate system transformation to convert the position of the ripe blueberry relative to the lidar 31 to the position relative to the blueberry picking mechanism 5. This allows the blueberry picking mechanism 5 to be driven to the position in front of the ripe blueberry.

[0060] After determining the position of the ripe blueberries relative to the blueberry picking mechanism 5, the automated blueberry harvesting robot moves to a position about half a meter in front of the blueberries via the moving mechanism 1. At this time, the multi-axis robotic arm 4 is activated, driving the end-effector blueberry picking mechanism 5 to a position closer to the blueberries. The moving motor then starts to rotate. The moving motor is bolted to the outside of the planetary gear system's carriage 11. The rotating shaft of the moving motor is directly connected to the internal gear of the planetary gear system. The rotating shaft of the moving motor drives the drive wheel to rotate. The drive wheel closely cooperates with the three planetary driven wheels around it. In this way, multiple fixed driven wheels can evenly distribute the load, reducing the load on each driven wheel, improving the load-bearing capacity and lifespan of the driven wheel system, and reducing the wear and failure risk of individual driven wheels.

[0061] The driving gear is a spur gear with a module of 2 and 18 teeth, while the fixed driven gear is a spur gear with a module of 2 and 36 teeth. The fixed driven gear meshes with the driving gear 3-2 on one side and with the external gear 3-4 on the other, forming a driven gear train. The two fixed driven gears are each 120 degrees apart. The outer ring of the external gear is equipped with a bearing whose inner diameter matches that of the external gear, which can effectively reduce the friction between the internal gear and the planetary gear train's housing 11, improve the service life of the planetary gear train, and enhance operational stability.

[0062] The rotation of the driving wheel drives three fixed driven wheels to rotate, which in turn drives the gear ring 55 to rotate. The end of the blueberry picking mechanism 5 is fitted onto the gear ring 55. The rotation of the gear ring 55 causes the end of the blueberry picking mechanism 5 to rotate around the center of the planetary gear system, thus enabling the front end of the blueberry picking mechanism 5 to achieve a human-hand-like picking method. Compared to a conventional crank-rocker mechanism, the crank-rocker mechanism improved with a planetary gear system has a larger transmission ratio, higher torque, and smoother output, effectively reducing the vibration and impact of the blueberry picking mechanism 5, thereby reducing the damage rate during blueberry picking.

[0063] Blueberry picking agency 5. Specific process for picking blueberries as follows: Figure 4 As shown, the end comb 582 of the blueberry picking mechanism 5, driven by the gear system, performs an elliptical motion, and its motion trajectory is as follows. Figure 4 As shown, two arrows indicate the direction of movement of the blueberry picking mechanism 5. The toothed ring 55 fixed at the end of the blueberry picking mechanism 5 rotates at a constant speed, thereby driving the blueberry picking mechanism 5 to swing back and forth, thus exhibiting a quick-return characteristic. When the blueberry picking mechanism 5 moves to... Figure 4 At the bottom of the elliptical trajectory shown, the speed begins to increase and a rapid return characteristic begins to appear. During this process, the blueberry picking mechanism 5 partially wraps around the blueberry and retracts it. At the same time, due to the high speed and certain impact force, the blueberry is more likely to fall off.

[0064] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An automated blueberry harvesting robot, characterized in that, include: Mobile mechanism (1), Controller (2), which is fixed to the moving mechanism (1); Blueberry recognition mechanism (3), the blueberry recognition mechanism (3) includes a lidar (31) and a color camera (32), the lidar (31) is fixed on the controller (2); the color camera (32) is fixed on the lidar (31); A multi-axis robotic arm mechanism (4) is located on one side of the controller (2) and fixed on the moving mechanism (1); Blueberry picking mechanism (5), which is detachably connected to the execution end of the multi-axis robotic arm mechanism (4); the controller (2) is electrically connected to the lidar (31), the color camera (32), the multi-axis robotic arm mechanism (4) and the blueberry picking mechanism (5) respectively; The moving mechanism (1) includes a carriage (11), a moving motor, a driving pulley, a driven pulley, a track (12), and a battery. There are two moving motors, both of which are fixed inside the carriage (11). The output shafts of the two moving motors extend out of the opposite sides of the carriage (11). There are two driving pulleys located outside the carriage (11) and respectively fitted onto the output shafts of the two moving motors. There are multiple driven pulleys that are rotatably connected to the opposite outer sides of the carriage (11). There are two tracks (12) that are correspondingly and rotatably connected to the driving pulleys and multiple driven pulleys on both sides of the carriage (11). The battery is placed inside the carriage (11). The controller (2) is fixed on the carriage (11) and electrically connected to the moving motor and the battery. The multi-axis robotic arm mechanism (4) is fixed on the carriage (11). The blueberry picking mechanism (5) includes a fixed box (51), a picking motor (52), a drive wheel (53), a driven wheel (54), a wheel rim (55), a fixed rod (56), a moving rod (57), and a picking box (58). The fixed box (51) is detachably connected to the execution end of the multi-axis robotic arm mechanism (4). The interior of the fixed box (51) has a cavity arranged parallel to the height direction, and the longitudinal section of the cavity is circular. The side of the fixed box (51) away from the carriage (11) has a movable outlet hole communicating with the cavity. The picking motor (52) is fixed to the outer side of the fixed box (51), and its output shaft extends into the fixed box (51). The output shaft of the picking motor (52) is parallel to the axis of the cavity. The drive wheel (53) is located in the cavity and is sleeved on the output shaft of the picking motor (52). There are multiple driven wheels (54) arranged in a circular pattern. The planetary arrangement is distributed on the outer periphery of the driving wheel (53), and the multiple driven wheels (54) are rotatably connected to the cavity wall of the cavity and are meshed with the driving wheel (53); the wheel rim (55) is rotatably connected to the cavity along the circumference of the cavity, and the inner surface of the wheel rim (55) is fixed with gear teeth, and the gear teeth are meshed with the multiple driven wheels (54); the fixed rod (56) is arranged parallel to the fixed box (51), and one end of it is fixed to the outer side of the fixed box (51) corresponding to the side of the movable outlet hole, and the other end of it is fixed with a sliding sleeve; the moving rod (57) is arranged along the length direction of the fixed rod (56), and one end of it is fixed to the wheel rim (55), and the other end of it slides through the movable outlet hole and the sliding sleeve; the picking box (58) is fixed to the other end of the moving rod (57), and the other end of it away from the vehicle box (11) is provided with a picking hole.

2. The blueberry automatic harvesting robot according to claim 1, characterized in that, The controller (2) includes a control cabinet (21), an inertial sensor (22), and a main control computer (23) with an embedded LIO-SAM mapping algorithm. The bottom surface of the control cabinet (21) is fixed to the top surface of the vehicle body (11). The inertial sensor (22) is fixed inside the control cabinet (21). The main control computer (23) is fixed inside the control cabinet (21) and is electrically connected to the moving motor, the inertial sensor (22), the lidar (31), the color camera (32), the multi-axis robotic arm mechanism (4), and the blueberry picking mechanism (5). The lidar (31) is fixed on the vehicle body (11).

3. The blueberry automatic harvesting robot according to claim 1, characterized in that, The picking box (58) includes a baffle, a bottom plate, a top plate (581), comb teeth (582), and side plates (583). The baffle is vertically fixed to the other end of the moving rod (57). The bottom plate and the top plate (581) are fixed vertically to each other on the side of the baffle away from the vehicle body (11). There are multiple comb teeth (582), all arranged along the length direction of the moving rod (57). One end of each comb tooth (582) is fixed at a distance from the end of the bottom plate away from the baffle. There are two side plates (583), both arranged along the length direction of the moving rod (57). The two side plates (583) are fixed to the sides of the bottom plate and the top plate (581), and both side plates (583) are fixed to the adjacent comb teeth (582). The picking hole is the gap between the top plate (581), the multiple comb teeth (582), and the two side plates (583).

4. The blueberry automatic harvesting robot according to claim 3, characterized in that, The picking box (58) also includes a blueberry transport tube (59), the bottom surface of the base plate is inclined downward from the edge to the center and the lowest point of the base plate is provided with a blueberry transport hole; one end of the blueberry transport tube (59) is connected to the blueberry transport hole and the other end of the tube is inserted into the carriage (11).

5. The blueberry automatic harvesting robot according to claim 3, characterized in that, All of the comb teeth (582) are made of rubber.

6. The blueberry automatic harvesting robot according to claim 3, characterized in that, The diameter of each of the comb teeth (582) decreases sequentially in the direction away from the base plate.

Images