An agricultural robot based on artificial intelligence for picking and irrigation collaborative work

Abstract

The application discloses a kind of based on artificial intelligence's picking irrigation collaborative operation agricultural robot, including walking chassis, picking mechanism, irrigation mechanism, visual perception module, sensing module and control unit.Picking mechanism is connected to the front area of walking chassis, for carrying out fruit picking.Irrigation mechanism is connected to the rear area of walking chassis, for carrying out irrigation.The application can realize the function fusion of picking and irrigation on single equipment by integrating picking operation and irrigation operation on the same walking platform, avoid the efficiency loss caused by picking equipment and irrigation equipment in traditional scheme Time-sharing operation, greatly improve the work efficiency;According to the fruit maturity identification result and soil-crown multi-dimensional state information, the picking priority and irrigation urgency quantitative index are dynamically generated by fuzzy reasoning, the work timing distribution of picking and irrigation is self-adaptively adjusted, and the collaborative operation of picking and irrigation is realized.

Description

Technical Field

[0001] This invention belongs to the field of agricultural intelligent equipment technology, specifically relating to an agricultural robot for collaborative harvesting and irrigation operations based on artificial intelligence. Background Technology

[0002] With the structural shortage of agricultural labor becoming increasingly prominent, intelligent agricultural equipment with robotics technology at its core has become an important support for promoting agricultural modernization and is beginning to diversify.

[0003] Chinese patent CN121195719A discloses a multi-processing robot for modern agriculture. Using a tracked robot as its core mobile carrier, it integrates various operational modules such as a harvesting robotic arm and a mowing mechanism through a top-mounted frame. Leveraging the terrain adaptability of its tracked structure, it moves flexibly in the field, aiming to achieve integrated completion of multiple operations such as harvesting and mowing. However, this robot's multi-functional integration is a simple mechanical superposition design; the harvesting and mowing modules operate independently, lacking an intelligent collaborative mechanism based on crop data sharing. Furthermore, the robot's operational logic remains primarily focused on surface field operations, failing to cover the two most crucial aspects of facility agriculture: fruit harvesting and precision irrigation, thus limiting its applicability.

[0004] Therefore, in order to address the aforementioned technical issues, it is necessary to provide an agricultural robot based on artificial intelligence for collaborative harvesting and irrigation operations.

[0005] The information disclosed in this background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide an agricultural robot based on artificial intelligence that can coordinate harvesting and irrigation operations, which can solve the problem of the lack of intelligent coordination in existing agricultural robots.

[0007] To achieve the above objectives, a specific embodiment of the present invention provides the following technical solution: An artificial intelligence-based agricultural robot for collaborative harvesting and irrigation includes a walking chassis, a harvesting mechanism, an irrigation mechanism, a visual perception module, a sensing module, and a control unit. The harvesting mechanism is connected to the front of the walking chassis and is used for harvesting fruit. The irrigation mechanism is connected to the rear of the walking chassis and is used for irrigation. The visual perception module is connected to the upper front of the walking chassis and includes an RGB-D depth camera, a multispectral imaging sensor, and a near-infrared illumination unit. The sensing module is located at the bottom front of the walking chassis. The control unit is communicatively connected to the walking chassis, the harvesting mechanism, the irrigation mechanism, the visual perception module, and the sensing module. The control unit embeds a deep learning fruit ripeness recognition model and a harvesting-irrigation collaborative decision-making algorithm.

[0008] In one or more embodiments of the present invention, the harvesting mechanism includes a base, a free arm, and an end effector. The base is connected to the walking chassis. The free arm is connected to the base. The end effector is connected to the end of the free arm.

[0009] In one or more embodiments of the present invention, the end effector includes three sets of flexible envelope grippers and a fruit delamination force sensor. The three sets of flexible envelope grippers are arranged at 120° intervals on the circumference. Each set of flexible envelope grippers includes two opposing flexible silicone grippers. An array of flexible pressure sensing sheets is embedded on the inner surface of the flexible silicone grippers.

[0010] In one or more embodiments of the present invention, the irrigation mechanism includes a mounting plate, a pair of connecting mechanisms, a spray frame, and a root zone drip irrigation pipe. The mounting plate is detachably connected to the walking chassis. The pair of connecting mechanisms is fixedly connected to the mounting plate. The spray frame is fixedly connected to the pair of connecting mechanisms, and the spray frame is provided with a plurality of fan-shaped atomizing nozzles. The root zone drip irrigation pipe is fixedly connected to the lower side of the connecting mechanisms, and the pipe wall of the root zone drip irrigation pipe has a plurality of micropores evenly distributed.

[0011] In one or more embodiments of the present invention, the connecting mechanism includes a connecting rod, a contact plate, and a retaining ring. The connecting rod is disposed between the mounting plate and the spray frame. The contact plate is fixedly connected to the mounting plate and to the connecting rod. The retaining ring is fixedly connected to the end of the connecting rod away from the contact plate and is fixedly connected to the spray frame.

[0012] In one or more embodiments of the present invention, a water tank is provided on the walking chassis, a connecting pipe is connected between the water tank and the spray frame, a branch pipe is connected between the connecting pipe and the root zone drip irrigation pipe, a control valve group is provided on both the connecting pipe and the branch pipe, and a vehicle-mounted collection box is provided on the water tank.

[0013] In one or more embodiments of the present invention, the control valve group includes a high-pressure solenoid valve, a low-pressure regulating valve, a flow meter, and a water pressure sensor, for independently switching control between the high-pressure sprinkler mode of the jet frame and the low-pressure drip irrigation mode of the root zone drip irrigation pipe.

[0014] In one or more embodiments of the present invention, the sensing module includes a soil moisture sensor, a soil temperature sensor, a soil conductivity sensor, and an infrared canopy temperature sensor.

[0015] In one or more embodiments of the present invention, the control unit includes an AI computing platform, a communication module, and a software module. The software module embeds the deep learning fruit maturity recognition model and the harvesting-irrigation collaborative decision-making algorithm. The harvesting-irrigation collaborative decision-making algorithm is an expert decision-making algorithm based on multivariate fuzzy rules. It takes the fruit maturity classification result output by the deep learning fruit maturity recognition model, the normalized vegetation index calculated by the visual perception module, the soil volumetric water content and crop water stress index obtained by the sensing module as input variables, generates a harvesting priority index and an irrigation urgency index through fuzzy reasoning, and dynamically plans the time sequence allocation scheme of harvesting and irrigation operations based on the weighted comprehensive score of the two.

[0016] In one or more embodiments of the present invention, the specific decision rules of the harvesting-irrigation collaborative decision-making algorithm include: when the harvesting priority index is greater than the irrigation urgency index, the harvesting operation in the target area is executed first, and the irrigation intensity parameters are adjusted according to the average contact force data collected by the fruit abscission force sensor during the harvesting process after harvesting is completed; when the irrigation urgency index is greater than the harvesting priority index, the irrigation operation in the target area is executed first, and the image data of the visual perception module is updated immediately after irrigation is completed to re-evaluate the harvesting strategy; when the difference between the two indices is within the preset dual threshold range, the parallel collaborative mode of harvesting operation and irrigation operation is executed simultaneously, in which the harvesting mechanism and the irrigation mechanism are controlled independently of each other.

[0017] Compared with existing technologies, the present invention provides an artificial intelligence-based agricultural robot for collaborative harvesting and irrigation. By organically integrating harvesting and irrigation operations onto the same walking platform, it achieves functional fusion of harvesting and irrigation on a single device, avoiding the efficiency loss caused by the separate operation of harvesting and irrigation equipment in traditional solutions, and significantly improving operational efficiency. Based on the fruit maturity recognition results and multi-dimensional soil-canopy state information, it dynamically generates quantifiable indicators of harvesting priority and irrigation urgency through fuzzy reasoning, adaptively adjusting the time sequence allocation of harvesting and irrigation operations, thus realizing collaborative harvesting and irrigation operations. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a first-angle perspective view of an agricultural robot for harvesting and irrigation collaborative operation based on artificial intelligence, according to an embodiment of the present invention. Figure 2 This is a second-angle perspective view of an agricultural robot for harvesting and irrigation collaborative operation based on artificial intelligence, according to an embodiment of the present invention. Figure 3 for Figure 2 Schematic diagram of the structure at point A in the middle; Figure 4 This is a partial structural diagram of an artificial intelligence-based agricultural robot for collaborative harvesting and irrigation in one embodiment of the present invention. Figure 5 This is a structural block diagram of an artificial intelligence-based agricultural robot for collaborative harvesting and irrigation in one embodiment of the present invention; Figure 6 This is a structural block diagram of a sensing module in one embodiment of the present invention; Figure 7 This is a structural block diagram of the control unit in one embodiment of the present invention; Figure 8 This is a schematic diagram illustrating the working principle of an artificial intelligence-based agricultural robot for collaborative harvesting and irrigation in one embodiment of the present invention.

[0020] Explanation of key figure labels: 1-Walking chassis, 2-Harvesting mechanism, 201-Base, 202-Free arm, 203-End effector, 204-Vehicle collection box, 3-Irrigation mechanism, 301-Mounting plate, 302-Connecting mechanism, 3021-Connecting rod, 3022-Contact plate, 3023-Fixing ring, 303-Spraying frame, 304-Water tank, 305-Connecting pipe, 306-Root zone drip irrigation pipe, 307-Branch pipe. Detailed Implementation

[0021] To enable those skilled in the art to better understand the technical solutions in this disclosure, the technical solutions in the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this disclosure.

[0022] like Figures 1 to 8 As shown, an artificial intelligence-based agricultural robot for collaborative harvesting and irrigation in one embodiment of the present invention includes a walking chassis 1, a harvesting mechanism 2, an irrigation mechanism 3, a visual perception module, a sensing module, and a control unit.

[0023] The walking chassis 1 can adopt a tracked walking mechanism to adapt to the soft and uneven terrain that may occur in facility agriculture, ensuring the stability and passability of the robot during operation. Of course, under the condition of hard and flat ground, the walking chassis 1 can also be replaced with a wheeled structure, and the present invention does not specifically limit it.

[0024] In addition, an inertial navigation module is installed at the center bottom of the walking chassis 1. This module includes a nine-axis inertial measurement unit (IMU) and a high-precision GPS / BeiDou dual-mode positioning module. The IMU integrates a three-axis accelerometer, a three-axis gyroscope, and a three-axis magnetometer, and uses an extended Kalman filter algorithm for attitude calculation, achieving a static heading angle accuracy of less than 0.3° RMS. The GPS / BeiDou dual-mode positioning module supports RTK differential positioning, achieving a horizontal positioning accuracy better than 2cm in RTK fixed-solution mode. The inertial navigation module is connected to the control unit via a CAN bus, providing the robot with seamless indoor and outdoor continuous dead reckoning and path tracking capabilities.

[0025] like Figures 1 to 4 As shown, the picking mechanism 2 is connected to the front area of ​​the walking chassis 1 and is used to perform fruit picking tasks.

[0026] The harvesting mechanism 2 includes a base 201, which is fixedly mounted on the walking chassis 1. A multi-degree-of-freedom free arm 202, with its proximal end connected to the base 201, can perform extension, pitch, and rotation movements for flexible positioning in space. An end effector 203, connected to the end of the free arm 202, is the actuator that directly contacts the fruit and performs clamping and separating actions.

[0027] In this embodiment, the free arm 202 adopts a six-degree-of-freedom joint series configuration, and the rotation axes of the six joints are precisely defined and arranged according to the industrial robot standard DH parameter table.

[0028] Specifically, from base 201 upwards, the joints are as follows: Joint 1 (waist rotation joint, rotation angle range ±170°, rated speed 60° / s, servo motor peak torque 5.2 N·m), Joint 2 (shoulder pitch joint, pitch angle range -85° to +95°, rated speed 45° / s, servo motor peak torque 8.8 N·m), Joint 3 (elbow pitch joint, pitch angle range -135° to +80°, rated speed 40° / s, servo motor...). The peak torque is 4.6 N·m), joint four (wrist rotation joint, rotation angle range ±175°, rated speed 80° / s, servo motor peak torque 2.2 N·m), joint five (wrist pitch joint, pitch angle range ±110°, rated speed 90° / s, servo motor peak torque 1.6 N·m), and joint six (end-effector rotation joint, rotation angle range ±360°, rated speed 120° / s, servo motor peak torque 0.9 N·m).

[0029] In addition, each joint is equipped with an absolute encoder to provide absolute position feedback of the joint angle, with an encoder resolution of no less than 17 bits and an angle resolution better than 0.0027°. Each joint servo drive supports online switching between three operating modes: torque control, speed control, and position control.

[0030] Furthermore, a vehicle-mounted collection box 204 is also fixed on the chassis 1 for temporary storage of the harvested fruit.

[0031] like Figures 1 to 4 As shown, the end effector 203 mainly includes three sets of flexible enveloping grippers arranged at 120° intervals on the circumference, and a fruit delamination force sensor integrated at the base of the grippers.

[0032] Each flexible envelope gripper consists of two opposing flexible silicone fingers. The silicone material itself has excellent flexibility and biocompatibility, which can effectively cushion the gripping force.

[0033] Furthermore, each flexible silicone gripper finger has an array of flexible pressure sensor pads embedded on its inner surface. These sensors can sense the pressure distribution at each contact point on the fruit's surface in real time during gripping and feed the pressure data back to the control unit. The control unit dynamically adjusts the closing torque of the grippers in real time based on the pressure feedback, forming a closed-loop control of sensing and gripping to ensure that the fruit is grasped with minimal gripping force.

[0034] In addition, the fruit abscission force sensor is used to measure the ultimate force required to separate the fruit from the branch when twisted or pulled. The magnitude of this force reflects the maturity of the fruit, the degree of lignification of the stem, and the tightness of the connection with the fruit stalk. On the one hand, it can help determine the harvesting time, and on the other hand, it provides an important basis for adjusting irrigation intensity parameters in collaborative decision-making algorithms.

[0035] like Figures 1 to 4 As shown, the irrigation mechanism 3 is connected to the rear area of ​​the chassis 1 for precision irrigation. The irrigation mechanism 3 includes: a mounting plate 301, which is detachably connected to the rear end of the chassis 1 by fasteners such as bolts, so as to facilitate the replacement of irrigation components of different specifications according to the crop row spacing; a pair of oppositely arranged connecting mechanisms 302, which are fixedly connected to the mounting plate 301; a laterally extending spray frame 303, which is fixed by the connecting mechanisms 302; and at least one drip irrigation pipe 306, which is fixedly connected to the lower side of the connecting mechanisms 302.

[0036] The connecting mechanism 302 specifically consists of a connecting rod 3021, a contact plate 3022, and a retaining ring 3023. The contact plate 3022 is bolted to the side of the mounting plate 301. One end of the connecting rod 3021 is fixed to the contact plate 3022, and the other end is fitted with the retaining ring 3023. The retaining ring 3023 grips the pipe body of the spray frame 303, thus providing stable support for the spray frame 303. Simultaneously, a root zone irrigation pipe 306 is fixed to the lower side of the contact plate 3022.

[0037] Specifically, the spray frame 303 is uniformly equipped with multiple fan-shaped atomizing nozzles along the axial direction. The angle and spacing of these nozzles can be adjusted according to the width of the crop canopy. In high-pressure irrigation mode, high-pressure water from the water tank 304 is transported to the spray frame 303 through the connecting pipe 305, and forms fine droplets through the fan-shaped atomizing nozzles, spraying and covering the crop canopy to achieve the functions of foliar humidification and cooling and humidification.

[0038] In addition, the walls of the root zone drip irrigation pipe 306 are uniformly distributed with a large number of micron-sized micropores. In low-pressure drip irrigation mode, water enters the root zone drip irrigation pipe 306 under low pressure, and the water slowly seeps out from the micropores, precisely delivering it to the soil area around the crop roots, greatly reducing surface evaporation loss and achieving high-efficiency water saving.

[0039] In addition, independent control valve assemblies are installed on both the connecting pipe 305 and the branch pipe 307 leading to the root zone drip irrigation pipe 306. These control valve assemblies include a high-pressure solenoid valve, a low-pressure regulating valve, a flow meter, and a water pressure sensor. When the system determines that foliar irrigation is required, the high-pressure solenoid valve on the connecting pipe 305 opens, and the valve on the branch pipe 307 closes. The reverse is true when root drip irrigation is required. The flow meter and water pressure sensor provide real-time feedback on the operating status, achieving closed-loop precise water and fertilizer control.

[0040] like Figures 4 to 8 As shown, the visual perception module is installed on the front top of the walking chassis 1, and integrates an RGB-D depth camera, a multispectral imaging sensor and a near-infrared supplementary light unit.

[0041] The RGB-D depth camera is used to acquire the spatial three-dimensional coordinates and morphological information of the fruit. The multispectral imaging sensor is used to acquire multispectral images of the crop canopy to calculate crop growth status parameters such as the normalized difference vegetation index. The near-infrared illumination unit is used to provide an active light source for the camera at night or under conditions of insufficient natural light, ensuring all-weather visual perception capability.

[0042] like Figures 1 to 6 As shown, the sensing module is located at the bottom front end of the walking chassis 1, close to the ground. This sensing module is mainly used to collect physicochemical parameters of the soil and crop root zone in situ, providing a data basis for irrigation decisions.

[0043] Specifically, the sensing module includes a soil moisture sensor for directly measuring soil volumetric water content, a soil temperature sensor for monitoring root zone temperature, a soil conductivity sensor for assessing soil salinity and nutrient status, and an array-type infrared canopy temperature sensor for non-contact monitoring of crop canopy temperature. The difference between canopy temperature and atmospheric temperature can be used to calculate the crop water stress index, a key indicator for determining whether crops are experiencing water shortages.

[0044] In this embodiment, the sensing module is encapsulated in an IP67-rated waterproof and dustproof aluminum alloy housing. The bottom of the aluminum alloy housing is equipped with a quick-release plug-in mechanism, enabling tool-free quick disassembly and installation of the sensing module. The sensing module 6 communicates with the control unit via an RS485 bus.

[0045] like Figures 1 to 8 As shown, the control unit is fixed inside or above the walking chassis 1 and establishes communication connections with the drive system of the walking chassis 1, the servo motor of the picking mechanism 2, the solenoid valve group of the irrigation mechanism 3, the vision perception module, and the sensor module through industrial communication protocols such as CAN bus or Ethernet. The control unit embeds a deep learning fruit ripeness recognition model and a picking-irrigation collaborative decision-making algorithm, which is the core of the entire robot to achieve intelligent collaborative operation.

[0046] The control unit includes a high-performance AI computing platform and a 4G / 5G or Wi-Fi communication module in terms of hardware, and a deep learning fruit maturity recognition model and a harvesting-irrigation collaborative decision-making algorithm embedded in the software.

[0047] Specifically, the control software module consists of the following main software components: image acquisition and preprocessing component, deep learning fruit maturity recognition model inference component, multi-source sensor data fusion component, harvesting-irrigation collaborative decision-making algorithm component, timing coordination and control command generation component, and data recording and remote communication component.

[0048] The image acquisition and preprocessing component acquires RGB-D and multispectral images from the visual perception module, performing preprocessing such as white balance correction, distortion correction, denoising, and image registration. The deep learning fruit maturity recognition model inference component loads model weights quantized and optimized using TensorFlow Lite, performs forward inference on the preprocessed images, and outputs instance segmentation masks, bounding boxes, maturity classifications, and confidence scores. The multi-source sensor data fusion component unifies soil sensor data, canopy temperature data, visual recognition results, and robot state data into a global world coordinate system for spatiotemporal alignment and fusion. The harvesting-irrigation collaborative decision-making algorithm component performs fuzzy inference decision-making based on the fused multidimensional data. The timing coordination and control command generation component generates specific control command sequences for each execution module based on the decision results and distributes them via the CAN-FD bus. The data recording and remote communication component stores operational data in a local database and can synchronize with a cloud management platform via 4G / 5G / WiFi.

[0049] The deep learning-based fruit maturity recognition model uses RGB-D camera and spectral images as input, employing a convolutional neural network to detect fruits and classify their maturity in the images. The harvesting-irrigation collaborative decision-making algorithm is an expert decision-making algorithm based on multivariate fuzzy rules. It uses four input variables: fruit maturity classification results, normalized vegetation index calculated by the visual perception module, and soil volumetric water content and crop water stress index obtained by the sensing module. Through a pre-designed fuzzy membership function and inference rule base, the algorithm generates a harvesting priority index and an irrigation urgency index. Finally, based on the weighted comprehensive score of these two indices, a dynamic planning scheme for the time-series allocation of harvesting and irrigation operations is developed.

[0050] For example, when the picking priority index of a certain work area is higher than the irrigation urgency index, the robot prioritizes planning its path to that area to perform the picking task. During the picking process, the system records the average contact force collected by the fruit abscission force sensor on the end effector 203 in real time. After the picking in that area is completed, the system adjusts the subsequent irrigation intensity parameters in reverse based on this average contact force data: if the picking force is generally too high, it indicates that the fruit may be insufficient in water and the plants are slightly wilted, thus automatically increasing the irrigation amount in that area. Conversely, if the irrigation urgency index is currently significantly higher, the robot prioritizes sprinkler or drip irrigation. After irrigation is completed, the vision module immediately re-collects image data of the area, reassesses the canopy status and fruit condition, updates the picking priority, and then determines the next path. When the scores of the two indices are extremely close, and the difference falls within the preset dual threshold range, the system enters a parallel collaborative mode, where the picking mechanism 2 and the irrigation mechanism 3 are independently controlled and operate simultaneously to maximize work efficiency.

[0051] In practical use, the control unit plans the autonomous navigation path of the walking chassis 1 between planting rows based on pre-loaded greenhouse map information or an environmental grid map constructed in real time via the visual perception module. After the robot travels along the planned path to the target planting area, it collects canopy image data through the visual perception module and simultaneously obtains soil moisture content, temperature, and electrical conductivity data through the sensing module. The control unit inputs the canopy image data into a deep learning fruit maturity recognition model and outputs maturity classification results and spatial location coordinate estimates for various target crops. The harvesting-irrigation collaborative decision-making algorithm integrates maturity classification results, normalized vegetation index, soil moisture content, and water stress index to generate a harvesting priority index and an irrigation urgency index. Based on the difference between the harvesting priority index and the irrigation urgency index, it determines whether to enter the priority harvesting mode, priority irrigation mode, or conflict-free parallel collaborative mode. In the priority harvesting mode or conflict-free parallel mode, the harvesting mechanism 2 performs precise harvesting, dynamically adjusting the clamping force amplitude by using a fruit abscission force sensor to provide real-time feedback on the grasping force, and placing the harvested fruit into the on-board collection box 204. In priority irrigation mode or conflict-free parallel mode, irrigation mechanism 3 selects high-pressure sprinkler irrigation mode or low-pressure seepage irrigation mode according to irrigation needs, and opens the corresponding valves to perform precise irrigation. After irrigation is completed, the sensor module rechecks the soil moisture content. If it does not reach the target humidity range, a supplementary irrigation cycle is started.

[0052] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0053] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0054] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0055] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0056] It will be apparent to those skilled in the art that this disclosure is not limited to the details of the exemplary embodiments described above, and that this disclosure can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of this disclosure is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within this disclosure. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0057] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. An agricultural robot for collaborative harvesting and irrigation based on artificial intelligence, characterized in that, include: Walking chassis; A harvesting mechanism, connected to the front area of ​​the walking chassis, is used for harvesting fruit; An irrigation mechanism, connected to the rear area of ​​the chassis, is used for irrigation; The visual perception module, connected to the upper front end of the walking chassis, includes an RGB-D depth camera, a multispectral imaging sensor, and a near-infrared supplementary lighting unit; The sensing module is located at the bottom front end of the walking chassis; The control unit is communicatively connected to the walking chassis, the picking mechanism, the irrigation mechanism, the visual perception module, and the sensing module. The control unit embeds a deep learning fruit maturity recognition model and a picking-irrigation collaborative decision-making algorithm.

2. The agricultural robot for collaborative harvesting and irrigation based on artificial intelligence according to claim 1, characterized in that, The harvesting facilities include: The base is connected to the walking chassis; A free arm is attached to the base; An end effector is attached to the end of the free arm.

3. The agricultural robot for collaborative harvesting and irrigation based on artificial intelligence according to claim 2, characterized in that, The end effector includes three sets of flexible envelope grippers and a fruit delamination force sensor. The three sets of flexible envelope grippers are arranged at 120° intervals on the circumference. Each set of flexible envelope grippers includes two opposing flexible silicone grippers. The inner surface of the flexible silicone grippers is embedded with an array of flexible pressure sensing sheets.

4. The agricultural robot for collaborative harvesting and irrigation based on artificial intelligence according to claim 1, characterized in that, The irrigation system includes: Mounting plate, which can be detachably connected to the walking chassis; A pair of connecting mechanisms are fixedly connected to the mounting plate; A spray frame is fixedly connected to a pair of the connecting mechanisms, and the spray frame is provided with a plurality of fan-shaped atomizing nozzles; The root zone drip irrigation pipe is fixedly connected to the lower side of the connecting mechanism, and the pipe wall of the root zone drip irrigation pipe has multiple micropores evenly distributed.

5. The agricultural robot for collaborative harvesting and irrigation based on artificial intelligence according to claim 4, characterized in that, The connecting mechanism includes: A connecting rod is located between the mounting plate and the spray frame; The contact plate is fixedly connected to the mounting plate and to the connecting rod. A retaining ring is fixedly connected to the end of the connecting rod away from the contact plate, and the retaining ring is fixedly connected to the spray frame.

6. The agricultural robot for collaborative harvesting and irrigation based on artificial intelligence according to claim 4, characterized in that, The walking chassis is equipped with a water tank, and a connecting pipe connects the water tank to the spray frame. A branch pipe connects the connecting pipe to the root zone drip irrigation pipe. Control valve groups are provided on both the connecting pipe and the branch pipe. A vehicle-mounted collection box is provided on the water tank.

7. The agricultural robot for collaborative harvesting and irrigation based on artificial intelligence according to claim 6, characterized in that, The control valve group includes a high-pressure solenoid valve, a low-pressure regulating valve, a flow meter, and a water pressure sensor, which are used to achieve independent switching control between the high-pressure sprinkler mode of the jet frame and the low-pressure drip irrigation mode of the root zone drip irrigation pipe.

8. The agricultural robot for collaborative harvesting and irrigation based on artificial intelligence according to claim 1, characterized in that, The sensing module includes a soil moisture sensor, a soil temperature sensor, a soil conductivity sensor, and an infrared canopy temperature sensor.

9. An agricultural robot for collaborative harvesting and irrigation based on artificial intelligence as described in claim 3, characterized in that, The control unit includes an AI computing platform, a communication module, and a software module. The software module embeds the deep learning fruit maturity recognition model and the harvesting-irrigation collaborative decision-making algorithm. The harvesting-irrigation collaborative decision-making algorithm is an expert decision-making algorithm based on multivariate fuzzy rules. It takes the fruit maturity classification results output by the deep learning fruit maturity recognition model, the normalized vegetation index calculated by the visual perception module, the soil volumetric water content and crop water stress index obtained by the sensing module as input variables. It generates a harvesting priority index and an irrigation urgency index through fuzzy reasoning, and dynamically plans the time sequence allocation scheme of harvesting and irrigation operations based on the weighted comprehensive score of the two.

10. An agricultural robot for collaborative harvesting and irrigation based on artificial intelligence as described in claim 9, characterized in that, The specific decision rules of the harvesting-irrigation collaborative decision-making algorithm include: when the harvesting priority index is greater than the irrigation urgency index, the harvesting operation in the target area is executed first, and the irrigation intensity parameters are adjusted based on the average contact force data collected by the fruit abscission force sensor during the harvesting process after harvesting; when the irrigation urgency index is greater than the harvesting priority index, the irrigation operation in the target area is executed first, and the image data of the visual perception module is updated immediately after irrigation to re-evaluate the harvesting strategy; when the difference between the two indices is within the preset dual threshold range, the parallel collaborative mode of harvesting and irrigation operations is executed simultaneously, in which the harvesting mechanism and the irrigation mechanism are controlled independently of each other.

Images