Experimental positioning control of flexible arm using two-degrees-of-freedom controller

Summary form only given. Flexible robot arm control design should cope not only with the inherent problem of flexible arm systems such as spillover instability but also with problems related to friction disturbances of the driving mechanism which negatively affect the tracking accuracy of the arm. To overcome the spillover instability, H/sup /spl infin// robust control theory can be employed for designing a feedback control system. However, in feedback control there is a trade-off between robust stability and tracking performance. To overcome this trade-off a 2-DOF controller is required. This paper proposes 2 design methods for designing robust positioning control systems of flexible robot arms using 2-DOF controllers, and compares the performance of both of them through experiment. Both methods design feedback controllers based on H/sup /spl infin// control theory. The two methods are different in how to design feedforward controller. The first method designs the feedback controller in the frequency domain by the use of the inverse dynamics of the flexible arm which is in this case a minimum-phase system. The second method, in contrast, designs the feedforward controller in the time domain using the inverse dynamics of the arm which is a nonminimum phase system.