This paper introduces the application of learning control theory to the intelligent control of an electromechanical cam-follower system. Learning control has been shown to handle inverse kinematics and inverse dynamics problems very well. It is a technique that can be applied to systems that perform repetitive tasks in order to reduce the errors that occur between the actual output and the desired output. Here, learning control is applied to a dynamic system containing nonlinear kinematics elements such as a cam. The learning process is based on output errors alone. It does not require explicit knowledge of the physical system. The presence of viscous damping and Coulomb friction both simplifies and challenges the learning control technique in compensating for such highly nonlinear dissipative effects within the cam-follower system. Results are presented based on the simulations of the system as well as the experiment. This study shows that learning control is capable of compensating for nonlinear Coulomb friction, that frequency occurs in the joints of many real world mechanisms.

This content is only available via PDF.
You do not currently have access to this content.