University of Technology MSc Programmes in Engineering Electrical Engineering Department of Computer Science and Electrical Engineering Division of Signal Processing 2006:184 CIV - ISSN ISRN: LTU-EX--06/184--SE
57
0
0
Full text
(2)
(3)
(4)
(5) . ! " # $ % . $ & ' ( ' & ( ' ) . $ # . . . . $ * . .
(6)
(7) & ) " +, * & & -. # + (/ ( ( ( & 0/ 0 & . .
(8)
(9)
(10)
(11) . .
(12)
(13) . . 1( 2 2
(14) .
(15).
(16) + 3
(17) 2 . + $
(18) . 4
(19)
(20) 5. . 6 - 62 " 62 622 * % 6
(21) & 66 8 66 8 662 # 8 66
(22) & $ . !" # . . . . ! 9 ' : !2 ;' ( 8 !
(23) ;' ( !
(24) * !
(25) 2 * !
(26)
(27) * ' !6 & ( ' . . . . . . . . . . . . . . # 7 & 7 2 7
(28) & 72 & 8 72 9 8 . . . . . . . 7. . . . %
(29) . . . 6 6 ! 7 3 3 3 3. $. 5 5 2< 2< 2 2 2
(30). . 2! 2! 24
(31) <
(32) <
(33) <.
(34) 722 8 72
(35) # 8 726 & $ 7
(36) & ' ( .
(37) . & . . . . . . . . . . . . . . . . . . .
(38) 2
(39)
(40)
(41) !
(42) 3. $ . 4 6 42 - ( 62. . %. .
(43)
(44)
(45)
(46) .
(47). &$ 8 % 8 & = & ) %
(48) 8 $ & . ( > ? . & % 8 (
(49) 8 & $ & ) $ & @ -$
(50) & . & 8 & ) ) & 8 # ' - A . 8 ( $ & ( ( & . $ (
(51) - ( ? & ( ' ( .
(52)
(53)
(54)
(55) .
(56)
(57)
(58)
(59) .
(60)
(61) . . .
(62) . .
(63) . - B & . . . & ( C D # C-. D 8 -. ( . $ & 8 # % & 8 ? . 2.
(64) . . 2 ( ? '
(65) & . 6 ' ! ( ' 7 ' 3 - 4 . (.
(66).
(67)
(68)
(69) 552 1 "E F G
(70) 0
(71) P M, C D ) . P M B C D P MH C2D @ % C";D H C
(72) D P "; & "; ; . ( ; & I . @ & . "J 9 ( ( . ( A = ( &
(73) F2G K A 554 % & - ) * ) # ( ( # - F
(74) G & ; % ;
(75) $ & ? ? . > ( F6G & % % F!G 6.
(76) # $ @ $ & $ & - 8 ? $ & + F7 3 4 5 < 2G . ) " ; @ "; 1 "E F G
(77) . &( + F
(78) G % ( ? # . & ( . - ' !<L & @ + F7G & 2 ( ) # ( ( 8 . ( - ? 8 . !.
(79) . & %. @. F 6G - F !G 1 F 7G F 3G F 4G - F
(80) G. . + ? F6G F
(81) G F!G > F 6G F 5G F2<G. . F
(82) G. $ F!G 0 F2 G 0 F22G. - 8. F
(83) 2
(84) G & F! 26G > F2!G $ $ F27G . F23G F4G ? F5 <G. " . "$ F6G . F24G 9 % F
(85) G ;' E ) F 6G. & 2 B . 7.
(86)
(87)
(88)
(89)
(90) .
(91) . $ ' ( ' & ) & J & B. ' () A ( :. '
(92) A : ' # * + 9 ( . : # % . ? : % ( & . % & ( . ' ,- *
(93) & . ? % . " ( 8 & C
(94) D . . $ & ( . $ A ( & & 10◦ ' & @ ( 3.
(95) & ) & % % # . % . % 0 ( . 8 & . $ & $ 8 ( # -
(96) . -
(97) B # > % ;' $ $ M. $ A . . . & % $ % & .
(98) $ % . + $
(99) . F25G -
(100) 2 & . $ $ & . ? $ $? & . $ $ $ 4.
(101) ? ? & ? . & ? ( . ? . -
(102) 2B . + $
(103) . & ) . ±20◦ @ . & ) $ $ ( . & 0, 0015◦ ( ' ' <3 & . 2<<< << ' . 8 2<<<<<$ <<<<<< . 2$4 ±10 CzD Cx y D . << & ±σ σ ≈ 68L & ) C1σD & ' ≈ 1.3 !< ≈ 2.6 << & ! !< 2< << & ' ? 8 & . 8
(104) !< ? 6L ? 2<L ' 4<< & !<< . . . & . $
(105) & ) NOP$) # ' ) . x y z & & & & ( @ ) $ 1 B 5.
(106) 2.
(107) . & x y z &
(108)
(109) z = (x, y) ( . A $ ' ' CzD . x y & ( $ ( x y #. C2D $ ' $ @ @ ' & '$ 1 ' # % - . & ) ( 8 "#&+#1 & 8 -. ' . <.
(110)
(111)
(112) 2 ? & ) P = {p1 , p2 , ..., pn } M = {m1 , m2, ..., mn } . n1 % n2
(113) & P M P M A
(114) p1 p1 & ' @ "; & P M P & "; & ' . 6 $6
(115) 66 ' 8 1. 2. . !
(116)
(117) . & & M P & ' n1n2 # ' ( @ . ; . > $ &@ F
(118) 6G # & @ & ' .
(119) @ . .
(120). .
(121)
(122)
(123)
(124)
(125) . .
(126)
(127)
(128)
(129)
(130) .
(131)
(132) . - @ ' 2 ' - 6 '
(133) $ @ @. 2.
(134)
(135)
(136) .
(137)
(138)
(139)
(140) .
(141)
(142) . . . .
(143)
(144)
(145) .
(146)
(147)
(148)
(149) . !
(150)
(151)
(152)
(153) . - 6 B ' F
(154) 6G & @ & ' & ' ) F
(155) <G n log n n @ . $ "#&+#1 delaunayn @ - @ . ' ) n log n A ' ( @ n log n ( A @ # $ delaunayn $ @ "#&+#1 & dsearchn ( . ' n log n
(156).
(157) & dsearchn M P & M P & % . M P M P % . . " . & "; & ; @ "; & . ; F
(158) G -
(159) CM D & M % & % ( 2 % ' 62 622 # ) N P M P M B C6 D R 3 × 3 ' T
(160) E A E & ( ) [R, T] P M & @ 8 R T % B M = RP + T + E ,. min R, T. Σ2 =. @ " % . N i=1. mi − Rp i − T. 2. C62D. RRT = I. C6
(161) D. det(R) = 1 .. C66D.
(162) . & # F
(163) 2G @ C62D - 3 × 3 ' &
(164) C6 D # C62D P M # ) % B N ¯ = 1 m i m N i=1. ¯ mci = m i − m. 6. C6!D.
(165) N 1 p p¯ = N i=1 i. p ci = p i − p¯ ,. C67D. C62D min R. =. N i=1. Σ2 =. N. i=1. T. mci − Rp ci2. T. T. (m ci m ci + p ci pci − 2m ci Rp ci ) .. C63D. & % @ '@ % '@ &(R H) & ' H ' ) B H=. N i=1. T. pci m ci .. C64D. > H H = UΛVT ' Rˆ '@ B C65D. ˆ = VUT . R. # ' Rˆ Tˆ M P 1 ( B ˆ p¯ . ˆ =m ¯ − R T. C6 <D.
(166) . & % C62D 3 × 3 M % . & . θ T ' a = [ax ay az ] & % . ) q = [cos(θ/2) sin(θ/2)ax sin(θ/2)ay sin(θ/2)az ]T & (
(167) * % F
(168)
(169) G ( C63D min R. . Σ =. N i=1. T m ciRpci. !. =. N i=1. (q p˙ ci q∗) · m ˙ ci ,. C6 D.
(170) q % R q∗ = [q0 − q1 − q2 − q3 ]T ˙ ci % ) @ p˙ ci m Q ;% C6 D . Σ = qT K q K 4 × 4 ' C6
(171) D ; K S & C6 2D Sab = ⎡ ⎢ ⎢ ⎣. K=⎢. N i=1. C6 2D. pcia mcib. Sxx + Syy + Szz Syz − Szy Szx − Sxz Sxy − Syx Syz − Szy Sxx − Syy − Szz Sxy + Syx Szx + Sxz Szx − Sxz Sxy + Syx Syy − Sxx − Szz Syz + Szy Sxy − Syx Szx + Sxz Syz + Szy Szz − Sxx − Syy. ⎤ ⎥ ⎥ ⎥ ⎦. C6
(172) D. & % qˆ '@ C6 D ' K 9 qˆ % ' Rˆ B ⎡. ⎤. ˆ2 − q ˆ 3) ˆ3 − q ˆ2) ˆ 21 − q ˆ 22 − q ˆ 23 ) ˆ0q ˆ 0q (ˆ q20 + q 2(ˆ q1 q 2(ˆ q1 q ⎢ 2 2 2 2 ˆ ˆ1 + q ˆ 3) ˆ3 − q ˆ1) ⎥ ˆ0q ˆ1 + q ˆ2 − q ˆ3) ˆ 0q (ˆ q0 − q 2(ˆ q2 q q2 q R = ⎣ 2(ˆ ⎦ . 2 2 2 2 ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ 2(ˆ q3 q1 − q0 q2 ) 2(ˆ q3 q2 − q0 q1 ) (ˆ q0 − q1 − q2 + q3 ). C6 6D. 9 ' Rˆ Tˆ C6 <D. . # . ˆ T ˆ . # ' R, & ( . "; $) . & . & & & ( ( # . $ A . 7.
(173) . $ . . $ ( ( C 2D . 8 .
(174) . . 8 8 ; # $) δmax . ; δmax & $$ δmax 9 δmax ( δmax # & $ δmax 8.
(175)
(176) . F23G P 8 δmax & & ( B. # 8 . μ < D δmax = μ + 3σ μ < 3D δmax = μ + 2σ μ < 6D δmax = μ + σ δmax = ξ. !! " . !! " !! " . . !! " . D ' & μ σ . & ξ ' .
(177)
(178)
(179) . & $ + F 2G & 8 % ' 7<L $ 3.
(180) & $ 1 ci ei - 62 . - 62B ) & % ci ei C6 !D C6 7D
(181). ci =. T mi. m . i (Rpi + T) mi . 2. Rpi + T − Rpi + T . ei = mi − Rpi − T. C6 !D C6 7D. + B |ei − eμ | < |ci − cμ| < κcσ A eμ cμ eσ cσ ei ci & κ ' + κ ∈ [1.0, 1.5] κ 8 ( κ . κeσ. 4.
(182)
(183)
(184)
(185) !
(186) "
(187) %. &' ( ). . ? . ' ( ? 8 & ? F6G & . 8 1 ? A . 8 ? 8 ( ? - ? A ? . ( ' * C D . & ) & 8 ) ( ( ) . %. *( ' +
(188) ,$. & ) ( ' 8 -. * J & ) $) $) 5.
(189) ( . & $ . ' ( . %. *( '
(190) . & ( ' C
(191) 22D ( # $( 2 $ ( & @ m × n ' (i, j) & ' (i, j) & ) & (i, j) # (i, j) (i, j + 1) & ' !
(192) . & ' ( & ( 8 @ ) 7 × 3 ' & ' C! D s(i, j) =. . . . m2 n2 1 (z(x, y) − z¯)2 , (m · n) − 1 y=m1 x=n1. . . . m m m1 = i − , m2 = i + , 2 2. . C! D . n n n1 = j − , n2 = j + 2 2. z¯ =. m2 n2 1 z(x, y) . (m · n) y=m1 x=n1. & ' @ & ' & ' ' ( ( & J ' ( 2<.
(193) ' 8 @ ( & ( @ '$ $' ( & ) ' $ * ' 8 !
(194) . 1 8J & . ( . . . & x y z # ) !
(195) C! D ( & ( '$ $' ) ( & ) !
(196) "
(197) . #
(198) . & ( ' & ) 8 ) J z$ & ) ( & # % . . & Slim & - p & ( pJ . p z$ 9 p & & ( & Slim % . . . & . % & . 8 Slim - $) ) 2.
(199) . ) & ss r α @ - ! . - ! B & * C!2D ss α tan = 2 2 r ss . C!2D r & . 9 α ( O r1 & ! A < <<<4 & ) 10 × 10 < . (10 · 0.008)2 = 0.0064 2 & r ) CD CD . < <<<4 << 2< << 7 2! !< <<6 6 << <<4 2<< < 7 <2! 2. & ! B $ . . O r1 Slim ' # Slim C!
(200) D 2. 50000 + 80 , r > 0 r2. C!
(201) D & 9 & ! <! 2 m2 . << & & & Cx y D ±7 !< ±10 << C1σD & ±10 !< << C1σD * . ( . Slim =. 22.
(202) & & . C!6D C!6D & ' ( & ) ' % ( 8 0 " # ? . # . @ y (x − z) y x − z A ) & ( ) & J . J 8 & 7 = 0.0004 · r + 0.02. %. # ' (
(203) . & & ) 8 . < 8 5< . . $ & - !2. .
(204) . - !2B & E ' ) @ 7 × 3 ' < - !
(205) & ( ' ( 2
(206).
(207) - !
(208) B E ' 9 ) ( ) & ) ) & @ 7 × 3 ' <<<! & - !6. - !6B E ' . & ( ' ' C!
(209) D C!6D - !! & 8 ( . # ( & . - !!B E ' ' 26.
(210)
(211) #
(212)
(213) -. ., . $
(214) . & ' & ) V = {V1, V2} & I . - 7 . . V1 . . V2 . - 7 B & & Fa = {Fa , Fa } Fb = {Fb , Fb } . $ & - 72 7
(215) & ( . ( & 1. 2. 2!. 1. 2.
(216) . Fa1 . . Fa2 . - 72B Fa & ( . . Fb1 . . Fb2 . - 7
(217) B Fb & (
(218) 2 - V ; <!$ & 10◦ ' & - 76. 27.
(219) - 76B V Fa Fb ; 2$
(220) @
(221) Fa Fb ( @ - 7! 77 2 Cx zD
(222)
(223) ; ( ( # ( (
(224) 2 ( ( .
(225) "
(226)
(227) .
(228)
(229)
(230) # .
(231) "
(232)
(233) . - 7!B Fa 23.
(234)
(235) "
(236)
(237) .
(238)
(239)
(240) # .
(241) "
(242)
(243) . - 77B Fb 72 < & ( Fa Fb ( & delaunayn "#&+#1 & "#&+#1 @ delaunayn % V A . $ & 7 & . ) & !
(244) 6 > CD CD V1 3425 <<4 < V2 4<72 <<4 < Fa 757<
(245) <63 56 Fa 757<
(246) <63 5
(247) Fb 757<
(248) <67 52 Fb 757<
(249) <64 5! 1 2. 1 2. & 7 B $ % "
(250) . & ( % - 73 . V 2 24.
(251) 8 CD = 0.01 D - 73CD - 73CD . x − z & I .
(252). V.
(253) .
(254)
(255)
(256)
(257)
(258)
(259)
(260)
(261). .
(262)
(263) #
(264) . V2 . . V1. - 73B M V 2 8 CD = 0.01 D - 74 Fa Fb
(265) 8 CD = 0.3 D & ) ( Fa Fb .
(266)
(267)
(268) "
(269). Fa1.
(270) . Fa2 . Fa . .
(271)
(272)
(273) "
(274) . Fb2 . Fb1. Fb . .
(275) . - 74B M Fa Fb
(276) 8 CD = 0.3 D 25.
(277) $ . ; . @ & Rˆ Tˆ . & √ . & NCc MSE c # 6 % √ NCc MSE c ; . & ) √ NCc . & MSE √ c 9 MSE c <<< L & √ ( ( MSE c 72 7
(278) √ & ) MSE c NCc & 8 ' # . # ' # & "#&+#1 ( % N. 24 0A@
(279) 01 #". -. # + /
(280) 0 $
(281) . $ & . & ) 8 # % $ & - 75 . x − z - 7 < 7 ( Fa Fb & ( x − z √ & . & 72 A MSE a ; NCa .
(282) <.
(283) V Fa Fb. √. MSE c. CD <<22
(284) <2!!5 <2744. NCc. √.
(285) <7 434
(286) 4475. & CD 4<7
(287) 63
(288) 7 6 377 !2 55 624
(289) ! 3
(290) 5. MSE a. NCa. CD <2
(291) 4< <375<
(292) 534. & 72B 8 .
(293). V.
(294) .
(295)
(296)
(297)
(298)
(299)
(300)
(301)
(302)
(303)
(304) # . V2 . . V1.
(305) . - 75B M V 8 & '$@ V1 V2 .
(306)
(307)
(308)
(309). Fa .
(310)
(311)
(312) "
(313). Fa1.
(314) . Fa2 . Fa . - 7 <B M Fa 8
(315). .
(316)
(317)
(318)
(319)
(320). Fb .
(321)
(322)
(323) "
(324). Fb . .
(325) . Fb2 . Fb1. - 7 B M Fb 8 $
Related documents
However, using position control in a large exible structure as a crane puts high demands on the control system.. The behavior of the crane also vary a great deal depending on
We have shown how an approximate LQG regula- tor, designed using a linear model, can be used to control an hydraulic actuator with a exible me- chanical load. The control system
Secondly, stability margins in the Nyquist plot are de- ned in terms of a novel clover like region rather than the conventional circle.. The clover region maps to a linearly
In the case the channel is modeled as a known tapped-delay line (nite impulse response lter) and the input has a nite number of possible values, the Viterbi algorithm pro- vides
However, employing the inherent structure of some important problems, such as H 1 and gain-scheduling synthesis, convexity can be recovered and the existence of a controller K can
In Figure 9 the change in resonance frequency, the response of estimated resonance frequency, and the inverse of the identication time constant is shown..
However, since the hinging hyperplane method only uses single-knot hinge functions at each step, where the tting would be as shown by the gure on the right, multiple hinge
A phenomenon that often occurs in the segmentation process is the spurious jumping in the state estimate of a hidden Markov model (HMM) when more states than needed are used..
