Page 105 - Autonomous Mobile Robots
P. 105

88                                     Autonomous Mobile Robots

                                The power reflected from the second feature, P REF2 is given by


                                                            2 ˆ               ])
                                                       P t GA I ϒ R 2  (1 −[ ˆ ϒ R 1  + ˆ ϒ a 1
                                               P REF2 =                   2               (2.31)
                                                                        ˆ
                                                           (4π) R (R 2 − R 1 )
                                                               2 ˆ 2 ˆ
                                                                 1
                                The power then transmitted back to the first feature from the second feature is
                                given by

                                                             3 ˆ
                                                        P t GA I ϒ R 2  (1 −[ ˆ ϒ R 1  + ˆ ϒ a 1 ])
                                               P INC21 =                                  (2.32)
                                                                           4
                                                                3 ˆ 2 ˆ
                                                                        ˆ
                                                            (4π) R (R 2 − R 1 )
                                                                  1
                                The power, P INC21 then passes through feature 1 and is given by
                                                                             ])           (2.33)
                                                 P TRAN21 = P INC21 (1 −[ ˆ ϒ R 1  + ˆ ϒ a 1

                                The power returned from the second feature is then P ˆ    = P TRAN21 A e /
                                                                           TRAN21
                                    ˆ 2
                                (4πR )
                                     1
                                                            3  ˆ                 2
                                                       P t GA I A e ϒ R 2  (1 −[ ˆ ϒ R 1  + ˆ ϒ a 1 ])
                                             P ˆ     =                                    (2.34)
                                              TRAN21                       4
                                                                         ˆ
                                                                4 ˆ 4 ˆ
                                                            (4π) R (R 2 − R 1 )
                                                                  1
                                   In general, the predicted power from the ith feature can be written as
                                                    KA I  (2i−1) ˆ ϒ R i (k + 1|k)
                                   ˆ
                                   P   TRANi1 (k + 1|k) =
                                                          (4π) 2i
                                                        i−1                           2
                                                        j=0  [1 −( ˆ ϒ R j  (k+1|k)+ ˆ ϒ a j (k + 1|k))]
                                                    ×                                     (2.35)
                                                          i−1  ˆ           ˆ        4
                                                          j=0 (R j+1 (k + 1|k) − R j (k + 1|k))
                                                         2                  ˆ
                                where K = P t GA e , A e = Gλ /4π, ˆ ϒ R 0  = ˆ ϒ a 0  = R 0 = 0 and, for the ith
                                       ˆ
                                feature, R i is related to the augmented state by Equation (2.25).
                                   Equation (2.25), Equation (2.26), and Equation (2.35) between them com-
                                prise the observation. In order to generate realistic predictions of the range bins,
                                knowledgeofthepowerandrangenoisedistributionsisnecessary. Thishasbeen
                                studied extensively in previous work, and can be found in Reference 15.
                                   The range and power noise are experimentally obtained [15]. The noise
                                in range is the phase noise, which is obtained by observing the range bins




                                 © 2006 by Taylor & Francis Group, LLC



                                 FRANKL: “dk6033_c002” — 2006/3/31 — 17:29 — page 88 — #48
   100   101   102   103   104   105   106   107   108   109   110