Page 39 - Adaptive Identification and Control of Uncertain Systems with Nonsmooth Dynamics
P. 39

30   Adaptive Identification and Control of Uncertain Systems with Non-smooth Dynamics


                           Substitute (2.37)–(2.41)into(2.43), and we have


                                   2
                                                              ˙
                          ˙ V  =−ks + σ 0 ˜z 0s − σ 1  |ω|  ˜ z 1s +  1  σ 0 ˜z 0 ˆ z 0 − ω −  |ω|  z 0  +  1  σ 1 ˜z 1
                                               g(ω)    k 0             g(ω)     k 1

                                 ˙        |ω|      γ σ 0  2  γ σ 1  2  γ ζ  ˜ 2  γ J 2  γ T ˜ 2
                                                                         ˜
                                 ˆ z 1 − ω −  z 1  −  ˜ σ −  ˜ σ −  ζ − J −    T
                                          g(ω)     k σ 0  0  k σ 1  1  k ζ  k J  k T  d
                                   2  σ 0 |ω|  2  σ 1 |ω|  2  γ σ 0  2  γ σ 1  2  γ ζ  ˜ 2  γ J 2  γ T ˜ 2
                                                                              ˜
                             =−ks −        ˜ z −    ˜ z −  ˜ σ −  ˜ σ −  ζ − J −     T
                                      k 0 g(ω) 0  k 1 g(ω) 1  0  k σ 1  1  k ζ  k J  k T  d
                                                        k σ 0
                                  1     ˙         |ω|
                                + σ 0 ˜z 0 ˆ z 0 − ω −  ˆ z 0 − k 0s
                                 k 0             g(ω)

                                  1     ˙         |ω|      |ω|
                                + σ 1 ˜z 1 ˆ z 1 − ω −  ˆ z 1 + k 1  s              (2.44)
                                 k 1             g(ω)      g(ω)
                           Then, substituting (2.28)into(2.44)yields
                                      σ 0 |ω|    σ 1 |ω|    γ σ 0  γ σ 1   γ ζ   γ J 2
                                             2
                                                                       2
                                                                              2
                                                                2
                                  2
                                                        2
                         ˙ V  =−ks −        ˜ z −      ˜ z −   ˜ σ −  ˜ σ −  ζ ˜ −  ˜ J
                                      k 0 g(ω)  0  k 1 g(ω)  1  k σ 0  0  k σ 1  1  k ζ  k J
                                 γ T
                                    ˜ 2
                               −    T ≤ 0                                           (2.45)
                                     d
                                 k T
                           From (2.45), we can obtain
                                                     ˙ V ≤−ks 2                     (2.46)
                           Integrating both sides of (2.46)from0to t,wehave
                                                                t
                                                                 2
                                              V(t) − V(0) ≤−k   s dτ                (2.47)
                                                              0
                        such that
                                                         t
                                                         2
                                              V(t) + k   s dτ ≤ V(0)                (2.48)
                                                       0
                                                t 2
                           Consider the facts k  s dτ ≥ 0, and V(t) ≥ 0from (2.42), then one can
                                              0
                        verify that
                                                      t
                                                      2
                                                 k    s dτ ≤ V(0)                   (2.49)
                                                    0
                        and
                                                   V(t) ≤ V(0)                      (2.50)
                                                                          t 2
                           From (2.49)and (2.50), we can conclude that k  s dτ and V(t) are
                                                                        0
                        bounded, which further implies that s, ˜z 0, ˜z 1, ˜σ 0, ˜σ 1, ζ are bounded.
                                                                           ˜
   34   35   36   37   38   39   40   41   42   43   44