CN-122004193-A - Intelligent weeding robot for dry soil crops

CN122004193ACN 122004193 ACN122004193 ACN 122004193ACN-122004193-A

Abstract

The invention discloses an intelligent weeding robot for dry soil crops, and relates to the technical field of intelligent agricultural equipment. The robot aims at the precision and adaptability requirements of weeding operation under a dry soil complex environment, and discloses a solution for integrating intelligent vision and self-adaptive control of edges. The robot comprises a movable trolley, a weeding device, a weed identification control module and a power battery, wherein the control module automatically controls the movable trolley to walk forwards according to the planned path and the position information provided by the satellite navigation module, receives the images of the dry soil crops and the weeds shot by the camera in real time, intelligently identifies the weeds and the coordinate information of the crops relative to the movable trolley by a main control unit, and timely controls the laser weeding device or the mechanical weeding device to realize efficient weed removal. The invention solves the problems of low efficiency and high seedling injury rate of the traditional weeding mode, and realizes intelligent, environment-friendly and efficient weeding.

Inventors

ZHANG DABING
ZHANG JUNYI
LI XINYE
LIU JUNQING
ZHOU RUICHENG
Bai Yanran

Assignees

湘潭大学

Dates

Publication Date: 20260512
Application Date: 20260313

Claims (5)

1. The intelligent weeding robot for the dry soil crops is characterized by comprising a mobile trolley, a weeding device, a weed identification control module and a power battery, wherein the mobile trolley consists of a frame and travelling wheels, the travelling wheels are fixedly arranged at the bottom of the frame-shaped frame, the weeding device consists of a laser weeding device fixedly arranged in the middle of the frame-shaped frame and a mechanical weeding device at the tail, the weed identification control module comprises a camera, a satellite navigation module, a main control unit and a motor driver, the camera, the satellite navigation module and the motor driver are electrically connected with the main control unit, the power battery is used for providing electric energy required by the whole machine work, the main control unit automatically controls the mobile trolley to walk forwards and receive the dry soil crops and weed photographs shot by the camera in real time according to position information provided by a planned path and the satellite navigation module, and the main control unit intelligently identifies weeds and coordinate information of the crops relative to the mobile trolley and timely controls the laser weeding device or the mechanical weeding device to realize efficient weed removal.
2. The intelligent weeding robot for dry soil crops according to claim 1, wherein the mobile trolley adopts a two-wheel or four-wheel driving four-wheel structure, and if a two-wheel driving mode is adopted, the two driving wheels can be positioned at the front wheel or the rear wheel.
3. The intelligent weeding robot for dry soil crops according to claim 1, wherein the laser weeding device in the weeding device comprises a laser generator, a laser vibrating mirror and a laser vibrating mirror controller, the laser generator is connected with the laser vibrating mirror through an optical fiber, the laser generator and the laser vibrating mirror controller are electrically connected with a main control unit, and the laser vibrating mirror is electrically connected with the laser vibrating mirror controller.
4. The intelligent weeding robot for dry soil crops according to claim 1, wherein the tail mechanical weeding device comprises a transverse sliding table module, a vertical sliding table module and a weeding component, wherein the transverse sliding table module is fixed on the tail of the frame-shaped frame, the vertical sliding table module is fixedly arranged on a sliding block of the transverse sliding table module, the weeding component is fixedly arranged on the sliding block of the vertical sliding table module, the weeding component consists of a cutter motor, a motor mounting plate and a rotary blade, the cutter motor is fixed on the sliding block of the vertical sliding table module through the motor mounting plate, and the rotary blade is fixedly connected with an output shaft of the cutter motor.
5. The intelligent weeding robot for the dry soil crops, which is disclosed in claim 1, is characterized in that the main control unit is electrically connected with the driving motors of the travelling wheels, the transverse slipway module and the vertical slipway module through a plurality of motor drivers, and is electrically connected with the cutterhead motor through the motor drivers, and the main control unit is used for identifying weeds and crops through running a lightweight deep learning target detection model (for example YOLOv series model), and the model can be used for rapidly identifying weeds and crop targets in a complex farmland background and obtaining coordinate information of weeds and crop targets through offline training of a large number of field crops and weed images.

Description

Intelligent weeding robot for dry soil crops Technical Field The invention relates to the technical field of agricultural machinery and intelligent robots, in particular to an intelligent weeding robot which is suitable for dry soil crops (such as corn and soybean) and can automatically identify and remove weeds. Background In dry soil crop planting, weeds compete with crops for water, nutrients and light, severely affecting yield. Currently, the primary modes of weeding include manual weeding, mechanical weeding and chemical weeding. The manual weeding efficiency is low, the labor intensity is high, crops and weeds cannot be distinguished by traditional mechanical weeding (such as a cultivator), the crops are easy to damage, the soil is difficult to work in the later period of crop growth, and the problems of soil pollution, pesticide residues, enhanced weed resistance and the like can be caused by the large-scale use of chemical herbicides. With the development of computer vision and artificial intelligence technology, intelligent weeding robots become a new approach for solving the problems. Disclosure of Invention The invention aims to disclose an intelligent weeding robot for dry soil crops, which solves the problems of low efficiency, poor precision or insufficient environmental protection of the existing weeding mode. The invention combines an efficient edge Artificial Intelligence (AI) calculation and reliable mechanical actuating mechanism and a laser weeding device, and aims to realize real-time, accurate and automatic removal of dry farmland weeds. In order to achieve the above purpose, the invention adopts the following technical scheme: The intelligent weeding robot for the dry soil crops comprises a movable trolley, a weeding device, a weed identification control module and a power battery, wherein the movable trolley consists of a frame and walking wheels, the walking wheels are fixedly arranged at the bottom of the frame-shaped frame, the weeding device consists of a laser weeding device fixedly arranged in the middle of the frame-shaped frame and a mechanical weeding device at the tail, the weed identification and control module comprises a camera, a satellite navigation module, a main control unit and a motor driver, the camera, the satellite navigation module and the motor driver are electrically connected with the main control unit, the power battery is used for providing electric energy required by the whole machine operation, the main control unit automatically controls the movable trolley to walk forwards and receive the images of the dry soil crops and the weeds shot by the camera in real time according to position information provided by a planning path and the satellite navigation module, and the main control unit intelligently identifies the weeds and the coordinate information of the crops relative to the movable trolley and timely controls the laser weeding device or the mechanical weeding device to realize efficient weed removal. Specifically, the travelling bogie adopts a four-wheel structure with two wheels or four wheels driven, if a two-wheel driving mode is adopted, the two driving wheels can be positioned at the front wheels or the rear wheels, and the main control unit can independently control the travelling speed of each driving wheel so as to ensure that the travelling bogie walks according to a planned path. Specifically, the laser weeding device in the weeding device comprises a laser generator, a laser vibrating mirror and a laser vibrating mirror controller, wherein the laser generator is connected with the laser vibrating mirror through an optical fiber, the laser generator and the laser vibrating mirror controller are electrically connected with a main control unit, the laser vibrating mirror is electrically connected with the laser vibrating mirror controller, and the laser weeding device is favorable for high-efficiency cauterizing and killing fine weeds. Specifically, afterbody machinery weeding device includes horizontal slip table module, vertical slip table module and weeding part, and horizontal slip table module is fixed in frame afterbody, and vertical slip table module fixed mounting is on the slider of horizontal slip table module, weeding part fixed mounting is on the slider of vertical slip table module, weeding part comprises cutter head motor, motor mounting panel and rotary blade, and cutter head motor passes through the motor mounting panel to be fixed in on the slider of vertical slip table module, rotary blade and cutter head motor output shaft fixed connection, and machinery weeding device is favorable to the rotary-cut robustly weeds. The main control unit identifies weeds and crops through running a light deep learning target detection model (for example YOLOv series models), the model can quickly identify weeds and crop targets in a complex farmland background and give out coordinate information of the weeds and crop targets through offlin