About this Event
Mechanical Engineering (ME)
Ph.D Dissertation Defense
"Data-Driven Control of a Strawberry Harvesting Manipulator"
University of California, Merced
Abstract:The United States of America is among the top ten strawberry producers globally. In recent years, mature strawberries are mostly left unharvested, and the production of strawberries is decreasing due to labor shortages, rising labor costs, and growing imports from Mexico. Many researchers are exploring robotic technologies to rejuvenate the strawberry industry, hoping that harvesting will be less dependent on seasonal human pickers. This research starts with designing and fabricating a manipulator and an end-effector for a small cooperative strawberry harvesting robot. It is intended to resolve the emerging lack of an affordable workforce in strawberry production.
In the next step, an inverse kinematic controller is developed for trajectory tracking of the designed manipulator. The developed control scheme is purely data-driven and does not require prior knowledge of the robot kinematics. Moreover, it can adapt to the changes in the kinematics of the robot. For developing the controller, the kinematic model of the delta robot is estimated by using neural networks. Then, the trained neural networks are configured as a controller in the system. The parameters of the neural networks are updated while the robot follows a path to adaptively compensate for modeling uncertainties and external disturbances of the control system.
Next, a data-driven optimal tracking control scheme is proposed using neural networks. First, a new neural networks model consisting of two networks in parallel is proposed to approximate the nonlinear dynamic system. Then the obtained data-driven model is used to design the optimal tracking controller. The developed controller consists of two parts, the steady-state controller and the optimal feedback controller. The optimal feedback control is developed based on approximating the solution of the Hamilton-Jacobi-Bellman equation by neural networks.
Akram Gholami is a Ph.D. candidate in the department of mechanical engineering. Her work is focused on modeling, simulating, and controlling robots using data-driven algorithms and machine learning. She received her B.Sc. degree in mechanical engineering from K. N. Toosi University of Technology, Iran; and M.S. degree in mechanical engineering and M.Eng. degree in agricultural and biological engineering, concurrently, from the University of Florida.
0 people are interested in this event