ABSTRACT
Typically, machines are used for simple, repetitive tasks in structured environments isolated from humans. However, a surge in active research in the area of human machine interaction is seen. Smart exercise machines, steer-by-wire applications, bilateral teleoperated robots, rehabilitation robots, and human assist gantry cranes are among the multitude of application areas that drive this research. While a variety of machine configurations are available to facilitate different exercises, many configurations can be reduced to a user torque input to an actuated motor. With that in mind, the specific problem that will be addressed is the design of a next generation exercise machine controller for a single degree of freedom system. For brevity, the extremum seeking algorithm is simply presented along with a heuristic commentary on the internal workings of the algorithm as opposed to extensive mathematical efforts to prove convergence of the scheme.
