| dc.description.abstract | As the demand for organic food grows, the concern of herbicides usage rises, and the herbicide-resistant weed surges. Compared to chemical, mechanical weed control has become a favorite in agricultural production. Precision weeding is becoming widely studied with the rapid development of sensors and computer technology. However, most of the current precision mechanical weeding is limited to large tractors or semi-automation systems. We focus on small robots and design an intelligent weeding robot with robotic arms traveling between crop rows.
This thesis discusses the design of the vehicle platform, the robotic arm, and the metal brush actuators. In the vehicle platform section, the vehicle parameters were selected, analyzed, and verified based on the experience of off-road vehicle designs. Key components of the vehicle, such as the A-arm, upright, and chassis, were analyzed and verified in the order of geometry-dynamics-FEA. In the design of the robotic arm, the number of active parts is reduced by using the characteristics of the weeding process, and a suitable control algorithm is developed. The structural part of the robot arm uses linkage system kinematics and topology optimization analysis to improve the overall performance. The weeding actuator design section draws on widely used agricultural tools to design four different forms of brush weeders. This part also predicts the working depth curve of the weeding brush under different soil moisture conditions by existing theories and verifies the correctness of the theories through experiments. Finally, the best brush structure is evaluated and presented.
Each part of the design in this thesis follows a theory-simulation/experiment-verification design flow. Given the growing trends of small agricultural vehicles, this design can provide many references. | |
