Page 479 - Mechanical Engineers' Handbook (Volume 2)
P. 479

470   Closed-Loop Control System Analysis

                             The CLCE is
                                              s(s   2)(s   3)   K (s   1)(s   4)   0
                                                              1
                          and its roots are the closed-loop poles appropriate for the specific gain value K .
                                                                                        1
                             If the transfer functions are rational (i.e., are ratios of polynomials), then they can be
                          written as
                                                              m
                                                           K   i 1  (s   z )
                                                                      i
                                                    G(s)     n                                (62)
                                                             j 1  (s   p )
                                                                     j
                             When the poles of the transfer function G(s) are distinct, G(s) may be written in partial-
                          fraction form as
                                                       G(s)      A j                          (63)
                                                              n
                                                             j 1  S   p j
                          Hence
                                               g(t)   L [G(s)]   L      n  A j
                                                       1
                                                                 1
                                                                   j 1  s   p j
                                                        Ae  pt j                              (64)
                                                      n
                                                     j 1  j
                          Since a transfer function G(s)   L[output]/L[input], g(t) of of Eq. (64) is the response of
                          the system depicted by G(s) for a unit impulse  (t), since L[ (t)]   1.
                             The impulse response of a given system is key to its internal stability. The term system
                          here is applicable to any part of (or the whole) closed-loop system.
                             It should be noted that the zeros of the transfer function only affect the residues A .In
                                                                                               j
                          other words, the contribution from the corresponding transient term e  pt j  may or may not be
                          significant depending on the relative size of A . If, for instance, a zero  z is very close to
                                                                                    k
                                                              j
                          a pole  p , then the transient term Ae  pt j  would have a value close to zero for its residue
                                  l
                          A . As an example, consider the unit impulse response of the two systems:
                           l
                                                         1          1    1
                                             G (s)                                            (65)
                                               1
                                                    (s   1)(s   2)  s   1 s   2
                                                      (s   1.05)   0.05   0.95
                                             G (s)                                            (66)
                                               2
                                                    (s   1)(s   2)  s   1  s   2
                                        (t)   e  t    e  2t , and from Eq. (66), g (t)   0.05e  t    0.95e  2t .
                          From Eq. (65), g 1                          2
                             Note that the effect of the term e  t  has been modified from a residue of 1 in G to a
                                                                                             1
                          residue of 0.05 in G . This observation helps to reduce the order of a system when there are
                                         2
                          poles and zeros close together. In G (s), for example, little error would be introduced if the
                                                      2
                          zero at  1.05 and the pole at  1 are neglected and the transfer function is approximated
                          by
                                                                1
                                                        G (s)
                                                         2
                                                               s   2
                             From Eq. (64) it can be observed that the shape of the impulse response is determined
                          primarily by the pole locations. A sketch of several pole locations and corresponding impulse
                          responses is given in Fig. 22.
   474   475   476   477   478   479   480   481   482   483   484