Design And Key Technology Research Of A Potted Plant Maintenance Robot System

At present,intelligent robot technology has become an important component of the manufacturing chain in the intelligent industry,especially mobile handling robots,which are playing an increasingly important role.Although there are various types of mobile handling robots,most of them focus on the fields of industrial handling robots or handling robotic arms.Generally speaking,due to the characteristics of small size,large quantity,and small coverage of potted plants,hiring dedicated personnel to handle,care for,and manage potted plants will increase management costs and human resources to a certain extent.In response to the above issues,this master’s project has preliminarily designed a potted plant maintenance robot system based on RGB D cameras,and conducted research on its key technologies,with a view to providing reference solutions and supporting technologies for realizing the automatic handling and maintenance of small potted plants in the home and office environment,effectively saving the labor workload and costs of manual handling.The system includes an omnidirectional mobile robot for potted plant handling and an indoor small intelligent greenhouse box for potted plant maintenance.The task of the indoor small intelligent greenhouse box for potted plant maintenance is to perform specific wavelength LED lighting,water and fertilizer showers,and other maintenance operations on the potted plants at the corresponding maintenance station.The task of an omnidirectional mobile robot for potted plant transportation refers to the combination of an omnidirectional moving chassis and a robotic arm,using information provided by visual sensors to complete tasks such as potted plant target recognition,grasping and handling,and sensing and navigating unknown environments.This topic mainly focuses on the following aspects of research work:Firstly,an omnidirectional mobile robot system for potted plant transportation was designed and initially implemented,and its embedded control system and software control process were designed.Secondly,an algorithm based on RGB-D cameras and Open CV was designed to estimate the key parameters required for pot transportation,aiming at obtaining the parameters such as depth of field,pot height,center point position,and diameter required for pot transportation.The algorithm process includes highly robust color threshold segmentation,dilation and erosion,connected domain extraction,edge detection,edge point depth information fusion,and multiplication fitting to obtain the centroid position,horizontal height,and fitting circle radius of the edge to be grasped in the flower pot,supporting the completion of the potted plant grasping.Thirdly,aiming at the visual servo control problem of omnidirectional mobile robot for potted plant handling,a robot docking servo strategy based on concise twodimensional code image features is proposed,including a method for calculating the paw posture of omnidirectional mobile robot for potted plant handling and a method for calculating the docking posture of omnidirectional mobile robot for potted plant handling.The robot captures the pixel coordinates of the four positioning points of the two-dimensional code using the two-dimensional code image directly below the vacant work position of the maintenance greenhouse box,and uses the pixel coordinate values to complete posture parking adjustment to achieve accurate parking.Finally,a self positioning algorithm based on the fusion of visual sensors and encoders is developed for the positioning and navigation of omnidirectional mobile robots for potted plant transportation.Experiments show that the trajectory quality and positioning accuracy of the self positioning algorithm designed in this paper are superior to the ORB-SLAM 3 algorithm alone,with better posture estimation accuracy.The experimental part of the paper proves the effectiveness of the designed system and algorithm,and provides a solution with certain reference value for the future research and development of omnidirectional mobile robot for potted plant transportation.