應用DGPS 差分定位系統於無人飛機之縱向氣動參數估算之應用研究(I)

(1)

DGPS (I) The Development and Application of DGPS Technique in Estimating

Longitudinal Aerodynamic Parameters for RPV.

NSC 88-2612-E-006-001

87.8.1-88.7.31

!"#$%

& ' ( & ) * + , - . / 0 1 2DGPS3456)789:*+

,-./01;<=>? DGPS/EKF 7 )@ABCDEFGHI' ! JKLM$N&OPQB014

R#O&STUV%&

WJX(&4&B GPS YZ&ST Garmin SvyII ;XL-25 ' Ashtech Z-12H WP[&(ST\1'G/]&)4Garmin YZ&O^_WE_ DGPS `abcI

`dO Z-12 *+./e'fgh01 ijk+lHImnopDEFG@A Oqrb`ab4opDEFG@A stHuI DGPS qrb7v!?w xyz{|2EKF3O}J~A4

DGPS/EKF #I@A&$

_)&.Orb+W

"OopDEFG4

DGPSHopDEFG@

Abstract

This reasrch studies the development and application of differential Global

Positioning System (DGPS) technique for self-developed remotely piloted vehicle (RPV) and ultralight airplane. The main goals of this dissertation include to implement cost-effective DGPS for kinematic positioning, to demonstrate a new methodology for estimating the longitudinal aerodynamic parameters from flight test data, and to develop a suitable aerial-photography payload incorporated with the DGPS navigation data. Three types of RPV and an ultralight airplane are used for flight tests and data/image acquisition. The on-board payload contains the GPS receivers, such as the Garmin SvyII, XL-25, and Ashtech Z-12, and a set of digital camera/CCD camera. Both static and kinematic DGPS accuracies of Garmin receivers are compared with the precise DGPS solution and the optical automatic tracking system (OATS), to ensure the accuracy so as to estimate the longitudinal aerodynamic parameters of the airplanes tested. The longitudinal aerodynamic parameters are studied by means of the Extended Kalman Filtering (EKF) with the navigation velocities obtained from DGPS.

The DGPS/EKF-based methodology is then used to obtain the trimmed flight test data.

(2)

Keywords: DGPS, estimating longitudinal aerodynamic parameters

&OWoppDEFG

d]4JHDEFG@

A+74JHt+l4 I USAF Datcom2197837H Etkin[1] Schmidt[2] ¡)¢£¤

¥¦ Datcom §¨©ª)«

¬<®?¯°4±HG²4

UH³´8<Hµ7¶5O·¸DE FGOCW4H¹º]

»A|¼½8<4¾§½<

i@ADEFGu¿+ÀÁÂÃÄÅ

ÆÇÈÉ)»A|H%&(

&uÊË')41980 ÌÍHklein and Batterson[3] Rabi[4]n °Î GÏI&BÐrÑ;ÒÓÑ Ñ|i

A&DEFG4

$%&(&ODEFG

¿+ÔÕHÖ×RØÙ Pioneer UAV

£ÚÛ=Î·¸G²H³´8<

.½<OÜÝ4ÞßàHHo[5]u

Jáâ# GPS )*+./eiãä )DEåæcGO¡4

çè<= C/A éê DGPS q

01è !R)%&(

&OE_./"#HëìÊíî7ï ði@ñ*+./01O`b4Cò)

¢IóôõA&O/

ö;rb;÷øùúøbû

!JKNüie GPS Oýþ

~AûÇO]

?ûM&B·¸. ªO WQBû`bO DGPS OATS f

gh01jk./`bû"#:) DGPSqrbeHI EKF i@A&

OopDEFG4² J¨·HK DGPS/EKF DEFG@AcI GPS 7}J»A|HË$I) GPS/EKF O./DEFGO@A4

3.1 GPS

GPS ,-./017Jp?)%

é./Hµ~A

+7HJ;éê~AH7ýþ

~Aû±;B{]/~A4

# C/A éêýþ~A¼

./eHýþ)~A?°

õ?(3.1)HD\ !

)"#$*WYZ%OÇùú$*

&HI'?(3.1))ýþ)~A

?°7?(3.2)4

( Si u) PRi i

Ri C dt dt

P =ρ + ⋅ − +ε (3.1)

( ) ( ) ( )

[ ⁱ û ² ⁱ û ² ⁱ û ²]⁰^.⁵

Ri X X Y Y Z Z

P = − + − + −

(dtSi dtu) dioni dtropi i

C⋅ − + + +ε

+ _, _, (3.2)

3.1.2

O"#R&O()N

$ 5 ®*Hb+:$ 300 ®,H -.qrb2Vn;Ve;Vd3IË/

0ω^L_EL×P^LO[142q3(NED)4 qrb+I·56O/ö78 9e4$÷øWùúøuI NED4qrb+.:4

(3)

E L E d

e n

L CP

h e h

a e h

e a

V V V

V &

&

=

















− 

 +

−



 



 +

−

=

















= λ φ

φ φ φ ) cos sin 1 (

) sin 1 (

) 1 (

5 . 0 2 2

5 .1 2 2

2

(3.3)

3.2 !"#$%&'(EKF)*

;k<r)=|\1?Ç#

wxyz{|2Kalman Filter;KF3H

I>?@JFAB²( Normal initial condition ) X^*H?CE_?D+

JFEF( reference trajectory )4v

!?wxyz{|2Extended Kalman Filter;EKF3)*GuFEFèHJ

~AIJKLM4ìN@

)/öeeµqrb+¢7 EKF )~AFGH#I@A&O&.

Orb+2u;w3;NOøbW7 892θ;q3;opDEFG2C_L;C_D; C_m3;opDEqG2C_Lα;C_mα3Wq

/ö2X;Y;Z34.Jq/ö'¢7 W DGPS /öej`bO#4

3.3 +,-./01.2

R ) U V % &

2OS-60 ;Falcon-2;Falcon-33I'(

&2MXLII3HµrbP:$ 40 m/s 2QRSG M=0.153HbT:$

400 ®,4-.HUVÖ× Etkin[1] WXOYìb&ZòO[=E

I\87OLop=E

4I7]&O X-Z ¸7

O^_`44-"8aOH

bcop=Eµ]õ 1. 3+,1.: ( v=0; p=r=0; φφφφ=0; ΨΨΨ = Ψ

constant)

2. 456789(KP):;<=>:

?

3. 45@7AB XbCDE: (T=P/VT, Ty=Tz=0)

4. FGHIJKL 5. MNOP0Q.:

(Vr_G Vr_TAS Vr_Wind Vr_T

≈ +

= )

6. RSFTUVWX:Y<: (δδδδthrottle, δδδδelevator, δδδδrudder, δδδδaileron )

Z G[\]^_`

¯J7«? RPV A½)\¯4RPV O&B GPS 7 Svy II îu XL-25H

P[d7\1]&2ef DC-40H 42mg3H¯±Q7Kh&B i¯;j “PL2”7 DGPS `bj k' INS A½QB4DGPS ON ücI&kt?¼4$ RPV OQI GPS 2NMEA l?3QI CCD[d\8<Hu¾§ UHF î 1.2GHzO[m\n|¼\4¯U7

O½<;DGPS e;

ÇoDEFG@AOp¯4

abcde

#)UV GPS YZ&¼

^ _ W E _ ` b j k H µ S T Garmin Survey II2\7 Svy II3HXL-25 ' Ashtech Z-124$¶H'~A GVqr)æHSvy II XL-25 st .7%&(&O½<RO E_./#H(&cu XL-25 Z-12 E_/ö;rb'÷øO`bj kHI'¨©qrb¢7opDEF G@AO~A4

5.1 GPS fghijkl

v8<c(&$wxßyz

§O&./eHçèÎeË GPS

(4)

YZ&)I{÷*4+ljk&

9HZ-12 W XL-25 O 2D W 3D ]

/ö+U$*7 6.9 ®,' 9.76 ® ,H]rbO+U$*7 1.92 ®,

|}4Z-12 W Svy II O]/örb O+²$*+G7 11 ®,22D3H14.85

®23D3I' 5.7 ®,|}4-.HSvyII O C/A éêCkOÉ~2Ðµ.

/$*4

5.2 DGPS fghijkl

^_ DGPS `buID+

j1OH+l&HSvy II O GPS

^_/öWrb«+$*78H95%

O±¸U/ö$*+G7 8.3 m' 13.1 mHµ±¸WUO rb$*7 2 m/s I' 2.6 m/s4XL-25 YZ&-Ù 12 jÈ C/A é êHÂ"# L₁B{µ./

eCkN)É~HÂ 95%O±¸

'Uö Wrb O$* +G 7 9.3 mH14.1 mH0.02 m/s W 0.15 m/s4Z-12 O^_ DGPS `bÜÝH$ 95%O±

¸U/ö$*+G7 1.8 m 3.5 mHrbO±¸'Urb$

*+G7 1.5 cm/sH' 3.1 cm/s4¯W

¯+GjkWIx]O U/öWrb213O$*4Ö×

Delio# Svy II ¢7O

#HÖO Svy II `bW Delio O`b]

j4ÖO Ashtech Z-12 `b+jµ

¢¡O Ashtech î Trimble OÜÝ`

aHÂ=8 XL-25 ëìB{

®+rb`b4 mne

èWX DGPS EKF $&

OopDEFGO@H#O:

DGPS èJÌ)'A½R

f[t)`mb4

è±Ì)R'¼&O

«E_A½DEFG+lH

Datcom+ljkHÙ)

Ý4

o/pq

[1] Etkin, B., “Dynamics of Flight,” Wiley, New York, Chap. 4, 1959

[2] Schmidt, S.F., “Computational techniques in Kalman Filtering,” in Theory and Applications of Kalman Filtering, AGARDograph 139, NATO Advisory Group for Aerospace Research and Development, London, Feb. 1970.

[3] Klein, V. and Batterson, J.G.

“Aerodynamic Parameters Estimated from Flight and Wind Tunnel Data,”

Journal of Aircraft, Vol.23, No.4, pp.306-312., 1985

[4]Trankle, T.L. and Rabin, U.H.,

“Filtering Flight Data Prior to Aerodynamic System Identification,”

AIAA Atmospheric Flight Mechanics Conference, AIAA-83-2098, pp.1-9, 1983.

[5] Ho, C.S., "A Study on Estimating Car Drag Parameter with Using GPS Solutions," Proceedings of the 1997 National Technical Meeting for Institute of Navigation, Santa Monica, CA 90744, Jan. 14-16, 1997

(5)

¯J Photography for the OS-60, ¯± Constellations of Payload design Falcon-2/3 RPV at different maneuvers: for RPV and ultra-light airplane [PL1:

DGPS/Aerial-photography mission; PL2:

DGPS kinematic calibration and INS flight test mission]

DGPS Processin g LS Method

Multi-Coord. Transformation [X, Y, Z]WGS-84

[[φφ, λ, λ, H], H]geodet ic ell ipsoidgeodeti c elli psoid

[E, N]2O TM

Navigation Velocity Computation VN, VE, VH, Ψ, γ GPS

INS

GPS Ground fixed station P R correct ion

α αα α, ββββ

Lon gi tudin al Aerodynam ic Coefficient Estimation

EKF al gorithm EKF algorit hm

Observation Model

Obse rvati on Model Dynamic ModelDynam ic Model

t, VN, VE, VH

a.< u, w, Θ,q, C_L,C_D,C_M>

b. < u, w, Θ, q,C_L, C_D, C_M, CLα, CMα >

c.< u, w, Θ,q, C_L,C_D,C_M, C_Lα, C_Mα >wi th P= f(V)

Initial states

P,Q states

Block diagram of the longitudinal aerodynamic coefficients estimation flow chart

GPS Receiver S urvey II

DC- 40 Camer a Camera

6th P CM Demodulator

C/A Code Ant.

X L-25

GPS Receiver Z-12

Code & P hase Ant.

DC- 40 Camer a

R ate Gy ro

PL1 PL1

PL2 PL2 B-1

B-2 B-3

B-4 B-5

A-1 A-2

A-3

A-4

(6)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 RMS erros (meters)

Ashtech MS- XII(Clynch,PDOP=2.5) Ashtech P-XII(Lachapelle)

Trimble 4000DL(Clynch,PDOP=2.5) Trimble TANS(Crow,PDOP=1.8) As htech Z-12(Guan, PDOP=1.8) Garmin XL-25(Guan, PDOP=1.8) Garmin SvyII(Guan, PDOP=1.8) Garmin SvyII(Delio)

C/A DGPS 3D position Comparison

Comparison of the DGPS 3D position errors (1σ) with recent studies

0 0. 25 0. 5 0.75 1 1.25 1.5 1. 75

R MS spee d erros (mete r/ s) Trimble TANS(Cwow, PDOP=1. 8)

Tri mbl e 4000DL(C lynnch, PDOP=2. 5) Ashtec h Z -12(Gua n, PDOP=1. 8) Garmin XL-25(Gua n, PDOP=1. 8) Garmin SvyII(Gua n, PDOP=1. 8)

C/A DGPS 3D speed comparison

Comparison of the DGPS 3D speed errors (1σ) with recent studies