Intelligent weeding robot for dry soil crops

Abstract

The invention discloses an intelligent weeding robot for dry soil crops, and relates to the technical field of agricultural intelligent equipment. According to the robot, a solution integrating edge intelligent vision and self-adaptive control is disclosed for the requirements for accuracy and adaptability of weeding operation in a dry soil complex environment. The robot comprises a moving trolley, a weeding device, a weed recognition control module and a power battery, and the control module automatically controls the moving trolley to walk forwards according to a planned path and position information provided by a satellite navigation module and receives dry soil crop and weed photos shot by a camera in real time. The main control unit intelligently recognizes coordinate information of weeds and crops relative to the moving trolley and timely controls the laser weeding device or the mechanical weeding device to efficiently remove the weeds. The problems that a traditional weeding mode is low in efficiency and high in seedling damage rate are solved, and intelligent, environment-friendly and efficient weeding is achieved.

Description

Technical Field

[0001] This invention relates to the field of agricultural machinery and intelligent robot technology, specifically to an intelligent weeding robot suitable for dryland crops (such as corn and soybeans) that can automatically identify and remove weeds. Background Technology

[0002] In dryland crop cultivation, weeds compete with crops for water, nutrients, and sunlight, severely impacting yield. Currently, the main weeding methods include manual weeding, mechanical weeding, and chemical weeding. Manual weeding is inefficient and labor-intensive; traditional mechanical weeding (such as cultivators) cannot distinguish between crops and weeds, easily damaging crops and having difficulty operating in the later stages of crop growth; the extensive use of chemical herbicides can lead to soil pollution, pesticide residues, and increased herbicide resistance in weeds. With the development of computer vision and artificial intelligence technologies, intelligent weeding robots have emerged as a new approach to solving these problems. Summary of the Invention

[0003] The purpose of this invention is to disclose an intelligent weeding robot for dryland crops, addressing the problems of low efficiency, poor accuracy, or environmental unfriendliness in existing weeding methods. This invention combines efficient edge artificial intelligence (AI) computing with reliable mechanical actuators and laser weeding devices, aiming to achieve real-time, precise, and automated removal of weeds in dryland fields.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A smart weeding robot for dryland crops includes a mobile cart, a weeding device, a weed recognition and control module, and a power battery. The mobile cart consists of a frame and wheels, with the wheels fixedly mounted at the bottom of the frame. The weeding device comprises a laser weeding device fixedly mounted in the middle of the frame and a mechanical weeding device at the rear. The weed recognition and control module includes a camera, a satellite navigation module, a main control unit, and a motor driver, all electrically connected to the main control unit. The power battery provides the necessary power for the robot's operation. Based on a planned path and location information provided by the satellite navigation module, the main control unit automatically controls the mobile cart to move forward and receives real-time photos of dryland crops and weeds from the camera. The main control unit intelligently identifies the coordinates of the weeds and crops relative to the mobile cart and promptly controls the laser or mechanical weeding device to efficiently remove the weeds.

[0006] Specifically, the mobile vehicle adopts a two-wheel or four-wheel drive four-wheel structure. If a two-wheel drive mode is adopted, the two drive wheels can be located at the front wheel or rear wheel position. The main control unit can control the travel speed of each drive wheel individually to ensure that the mobile vehicle travels according to the planned path.

[0007] Specifically, the laser weeding device in the weeding device consists of a laser generator, a laser galvanometer, and a laser galvanometer controller. The laser generator and the laser galvanometer are connected by optical fiber. The laser generator and the laser galvanometer controller are electrically connected to the main control unit. The laser galvanometer is electrically connected to the laser galvanometer controller. The laser weeding device is conducive to efficiently burning and killing small weeds.

[0008] Specifically, the tail-end mechanical weeding device includes a horizontal slide module, a vertical slide module, and a weeding component. The horizontal slide module is fixed to the tail of the frame, and the vertical slide module is fixedly installed on the slider of the horizontal slide module. The weeding component is fixedly installed on the slider of the vertical slide module. The weeding component consists of a blade disc motor, a motor mounting plate, and rotating blades. The blade disc motor is fixed to the slider of the vertical slide module through the motor mounting plate, and the rotating blades are fixedly connected to the output shaft of the blade disc motor. The mechanical weeding device is beneficial for rotary cutting of thick weeds.

[0009] Specifically, the main control unit is electrically connected to the drive motors of the traveling wheels, the horizontal slide module, and the vertical slide module through multiple motor drivers. At the same time, the main control unit is electrically connected to the cutter head motor through the motor drivers. The main control unit identifies weeds and crops by running a lightweight deep learning object detection model (such as the YOLOv5 series model). This model has been trained offline on a large number of field crop and weed images and can quickly identify weed and crop targets in complex farmland backgrounds and provide their coordinate information. After being trained on a large number of field crop and weed images, the accuracy of weed and crop identification has been improved.

[0010] The working principle of this invention is as follows: the robot moves autonomously or semi-autonomously between crop rows. The main control unit receives photos of dryland crops and weeds taken by the camera in real time. The model trained in the main control unit intelligently identifies the coordinate information of the weeds and crops relative to the moving cart and controls the laser weeding device or mechanical weeding device in a timely manner to achieve efficient removal of weeds.

[0011] The beneficial effects of this invention are as follows:

[0012] 1. By using edge AI for real-time image recognition, it can accurately distinguish between crops and weeds and achieve targeted removal of individual weeds, avoiding the "accidental damage" problem of traditional mechanical weeding and improving the accuracy and efficiency of weeding.

[0013] 2. It integrates environmental perception, decision control, and mechanical execution functions, which can significantly reduce the intensity of manual labor.

[0014] 3. Weed control methods using purely physical or laser burning methods completely avoid the use of chemical herbicides, which is beneficial to protecting the ecological environment and ensuring the safety of agricultural products.

[0015] 4. The modular design allows the visual algorithm and execution mechanism to be adjusted and optimized according to different crops and field conditions. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the overall structure of an intelligent weeding machine according to an embodiment of the present invention.

[0017] In the attached diagram: 1-Frame; 2-Walking wheel; 3-Camera; 4-Satellite navigation module; 5-Main control unit; 6-Motor driver; 7-Laser generator; 8-Laser galvanometer; 9-Laser galvanometer controller; 10-Horizontal slide module; 11-Vertical slide module; 12-Cutter head motor; 13-Rotating blade; 14-Power battery. Detailed Implementation

[0018] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0019] like Figure 1 As shown, one embodiment of the present invention provides an intelligent weeding robot for dryland crops. The robot comprises four main parts: a mobile cart, a weeding device, a weed recognition and control module, and a power battery 14. The mobile cart consists of a frame 1 and wheels 2, which are fixedly installed at the bottom of the frame 1 for autonomous movement in the field. The weeding device includes a laser weeding device located in the middle of the frame 1 and a mechanical weeding device at the rear, allowing for flexible selection of laser or mechanical weeding based on weed type, crop location, and operational requirements. The weed recognition and control module includes a camera 3, a satellite navigation module 4, and a motor driver 6, electrically connected to a main control unit 5. During operation, the main control unit 5 automatically controls the mobile cart to move along the crop path based on a preset work path and real-time position information fed back by the satellite navigation module 4. Simultaneously, the camera 3 collects real-time images of dryland crops and weeds, transmitting them to the main control unit 5 for intelligent recognition to obtain the position coordinates of the weeds and crops relative to the robot. Based on this, the main control unit 5 controls the laser or mechanical weeding device in real-time to achieve precise and efficient weed removal.

[0020] In this embodiment, the mobile trolley adopts a four-wheel structure, with either two-wheel or four-wheel drive. If it is two-wheel drive, the driving wheels are set as the front wheels, and the rear wheels are the driven wheels, ensuring that the robot moves smoothly and steers flexibly in the field. The frame 1 adopts a frame-shaped frame, which has high structural strength, facilitates the installation of the middle and rear devices, and also facilitates movement in the field, reducing operational congestion.

[0021] In this embodiment, the laser generator 7 and the laser galvanometer 8 are connected via optical fiber to achieve stable laser energy transmission. The laser generator 7 and the laser galvanometer controller 9 are electrically connected to the main control unit 5, and the laser galvanometer 8 is electrically connected to the laser galvanometer controller 9. The main control unit 5 outputs control signals to the laser galvanometer controller 9, which drives the laser galvanometer 8 to deflect, enabling rapid scanning and precise positioning of the laser spot in the field plane, burning and killing small weeds without damaging crops.

[0022] In this embodiment, the horizontal slide module 10 is fixedly installed at the rear of the frame 1 to realize the horizontal movement of the weeding component. The vertical slide module 11 is fixedly installed on the slider of the horizontal slide module 10 to realize the vertical lifting and lowering adjustment of the weeding component. The weeding component is fixedly installed on the slider of the vertical slide module 11. The blade motor 12 drives the rotating blade 13 to rotate at high speed to chop and remove the roots of thick weeds.

[0023] In practical implementation, the invention works as follows: the main control unit 5 uses a built-in program to plan the operational path for the dryland crop planting area and stores the planned path in the main control unit 5. The satellite navigation module 4 collects the robot's position information in real time and transmits it to the main control unit 5 to complete its self-positioning and initial calibration. Based on the preset planned path and the real-time position feedback from the satellite navigation module 4, the main control unit 5 drives the walking wheels 2 to rotate via the motor driver 6, controlling the mobile trolley to move forward autonomously at a constant speed along the crop rows. During the robot's movement, the camera 3 installed at the front or middle of the frame 1 continuously captures field images and transmits them to the main control unit 5 in real time. The main control unit 5 processes and recognizes the images, identifying crops and weeds in the complex farmland background, and calculates the lateral coordinates and longitudinal distance of each weed relative to the robot. Based on the identified weed coordinates, weed size, and distance from the crops, the main control unit 5 intelligently selects the weeding method. For small weeds, the laser weeding device is activated first, while for large weeds, the mechanical weeding device at the rear is used for removal. The main control unit 5 receives feedback information from each module in real time and dynamically adjusts the walking speed, weeding timing, and actuator actions to ensure that the robot does not damage crops or miss weeds during its movement. After the robot completes the weeding operation on the entire dry land area according to the planned path, the satellite navigation module 4 reports the endpoint location information, the main control unit 5 controls the walking wheels 2 to stop rotating, the laser weeding device and the mechanical weeding device are powered off and reset in sequence, and the robot automatically stops, completing a full weeding operation.

Claims

1. An intelligent weeding robot for dryland crops, characterized in that: The system includes a mobile trolley, a weeding device, a weed identification and control module, and a power battery. The mobile trolley consists of a frame and wheels, with the wheels fixedly mounted to the bottom of the frame. The weeding device comprises a laser weeding device fixedly mounted in the middle of the frame and a mechanical weeding device at the rear. The weed identification and control module includes a camera, a satellite navigation module, a main control unit, and a motor driver. The camera, satellite navigation module, and motor driver are electrically connected to the main control unit. The power battery provides the electrical energy required for the entire machine to operate. Based on the planned path and the location information provided by the satellite navigation module, the main control unit automatically controls the mobile trolley to move forward and receives real-time photos of dryland crops and weeds taken by the camera. The main control unit intelligently identifies the coordinates of the weeds and crops relative to the mobile trolley and promptly controls the laser weeding device or mechanical weeding device to achieve efficient removal of weeds.

2. The intelligent weeding robot for dryland crops according to claim 1, characterized in that: The mobile vehicle adopts a two-wheel or four-wheel drive four-wheel structure. If a two-wheel drive mode is adopted, the two drive wheels can be located at the front wheel or rear wheel position.

3. The intelligent weeding robot for dryland crops according to claim 1, characterized in that: The laser weeding device in the weeding device consists of a laser generator, a laser galvanometer, and a laser galvanometer controller. The laser generator and the laser galvanometer are connected by optical fiber. The laser generator and the laser galvanometer controller are electrically connected to the main control unit. The laser galvanometer is electrically connected to the laser galvanometer controller.

4. The intelligent weeding robot for dryland crops according to claim 1, characterized in that: The tail-end mechanical weeding device includes a horizontal slide module, a vertical slide module, and a weeding component. The horizontal slide module is fixed to the tail of the frame, and the vertical slide module is fixedly installed on the slider of the horizontal slide module. The weeding component is fixedly installed on the slider of the vertical slide module. The weeding component consists of a cutter head motor, a motor mounting plate, and rotating blades. The cutter head motor is fixed to the slider of the vertical slide module through the motor mounting plate, and the rotating blades are fixedly connected to the output shaft of the cutter head motor.

5. The intelligent weeding robot for dryland crops according to claim 1, characterized in that: The main control unit is electrically connected to the drive motors of the traveling wheels, the horizontal slide module, and the vertical slide module through multiple motor drivers. At the same time, the main control unit is electrically connected to the cutter head motor through the motor drivers. The main control unit identifies weeds and crops by running a lightweight deep learning target detection model (such as the YOLOv5 series model). This model has been trained offline on a large number of field crop and weed images and can quickly identify weed and crop targets and obtain their coordinate information in complex farmland backgrounds.

Images