Page 202 -
P. 202

4.4 Computed-Torque Control                                  185

            subsequent examples. The use of such simulation software as SIMNON is
            quite similar.
              We discuss the implementation of digital robot arm controllers in Section
            4.5. Some detailed discussion on digital control, simulation, and DSP
            implementation of controllers is given in [Lewis 1992].


            4.4 Computed-Torque Control


            Through the years there have been proposed many sorts of robot control
            schemes. As it happens, most of them can be considered as special cases of the
            class of computed-torque controllers. Computed torque, at the same time, is a
            special application of feedback linearization of nonlinear systems, which has
            gained popularity in modern systems theory [Hunt et al. 1983], [Gilbert and
            Ha 1984]. In fact, one way to classify robot control schemes is to divide them
            as “computed-torque-like” or “noncomputed-torque-like.” Computed-torque-
            like controls appear in robust control, adaptive control, learning control, and
            so on.
              In the remainder of this chapter we explore this class of robot controllers,
            which includes such a broad range of designs. Computed-torque control allows
            us to conveniently derive very effective robot controllers, while providing a
            framework to bring together classical independent joint control and some
            modern design techniques, as well as set the stage for the rest of the book. A
            summary of the different computed-torque-like controllers is given at the end
            of the section in Table 4.4.1. We shall see that many digital robot controllers
            are also computed-torque-like controllers (Section 4.5).
            Derivation of Inner Feedforward Loop

            The robot arm dynamics are

                                                                       (4.4.1)

            or

                                                                       (4.4.2)


            with the joint variable             the control torque, and    d(t) a
            disturbance. If this equation includes motor actuator dynamics (Section 3.6),
            then  (t) is an input voltage.
              Suppose that a desired trajectory q d(t) has been selected for the arm motion,
            according to the discussion in Section 4.2. To ensure trajectory tracking by
            the joint variable, define an output or tracking error as




            Copyright © 2004 by Marcel Dekker, Inc.
   197   198   199   200   201   202   203   204   205   206   207