( )

Size: px

Start display at page:

Download "( )"

しょうこすえがら
7 years ago
Views:

1 NAIST-IS-MT

2 ( )

3 Carmen Tookit Pioneer-2,,,, NAIST-IS- MT , i

4 A study of map information for path planning which combine velocity and safety Yuki Arai Abstract Recently, the service robot that does the work of the guide, guard, the guide, and nursing in an environment the same as human is a topic. These service robots need the function that autonomous mobile when they provide service. Mapping,localization and path planning are necessary for the function of the autonomous mobile. A cost map expressing easiness of street in the map by path finding is used, and the path varies with the value of the cost map. Generally the generation of the cost map is given depending on the distance from an obstacle, but, as for the generation method, a too deep argument is not done. Therefore We suggested sonar map, speed map, environment change map as technique newly to allot the cost to a map in this study. I showed the effectiveness of the proposal technique with simulation and an actual machine with mobile robot application Carmen Tookit and mobile robot Pioneer-2. As a result, the speed map was able to achieve shortening of movement time on simulation. In addition, by the experiment in the true environment where I used an environmental variation map for, Robot avoid dangerous places and succeeded in an obstacle and the path planning that did not collide. Keywords: mobile robot, laser range sensor, sonar, cost map, pathplaning Master s Thesis, Department of Information Systems, Graduate School of Information Science, Nara Institute of Science and Technology, NAIST-IS-MT , February 5, ii

Mapping,localization and path planning are necessary for the function of the autonomous mobile.

5 Carmen iii

4.1.............. 15 2.5............................ 16 2.5.1............... 17 2.5.2.............. 17 2.6................................. 18 3. 19 3.1................................ 19 3.1.1 Carmen.

6 A. CARMEN Toolkit 48 A A A A A A A A A A iv

3............................ 49 A.4................................ 51 A.4.1.................. 51 A.4.2......................... 51 A.5.............................. 52 iv

7 1 EMIEW ICP Waverront lpn Pioneer CAD v

....................... 17 14........................... 18 15 Pioneer3................................ 20 16................................ 21 17 CAD..................... 23 18.............................. 25 19.

8 A B C D E carmen navigatorgui robotgui Pioneer A B C D E vi

............................... 52 1........................ 14 2 Pioneer-3........................... 19 3...................... 20 4................................ 26 5....................... 30 6.

9 [14] ALSOK [10] wakamaru[11], enon EMIEW Fig.1 [3] Fig. 1 EMIEW 1.2 1

10 SLAM (Simultaneous Localization And Mapping) [7] SALM Thrun [9] Besl [9] Thrun [8] A*[2] [6] A* A* [12] ( ) [13] 2

11 A Carmen 3

12 ˆ ˆ ˆ ICP (MCL)[8] [6] A*[2] [4] ˆ ˆ ˆ 4

13 Fig2 GPS SLAM(Simultaneous Localization and Mapping) SLAM 2 Carmen [1] ICP (Iterative Closet Point ) 2 5

Fig. 2 2 2 M k = C(S k, T ) (1) E(T ) = i+1 N(M i (Rs i

14 Fig M k = C(S k, T ) (1) E(T ) = i+1 N(M i (Rs i + T )) 2 min (2) M s N R T ICP 1. (Fig.3(a)) (Fig.3(b)) 6

15 2. (Fig.3(c)) 3. (b) (c) 2 2 (Fig.3(d)) Fig. 3 ICP SLAM ICP s i M i ICP R, t Fig.4 ICP Fig.5 7

16 Fig. 4 Fig. 5 ; Fig.4. 8

17 2.2.2 bel(x t ) (3) X t = x [1] t, x [2] t,... x [M] t (3) x t [m] t M M 1. X 0 2. u t Xt 1 X t X t 3. x [m] t w [m] t = p(z t x [m] t ) (4) z t brl(x t ) bel(x t ) bel(x t 1 ) 4. X t M w [m] t 9

18 2 5 (Monte Carlo Localization MCL) (Monte Carlo Localization MCL) Fig.6 MCL p(z x) x z Fig.6(a) MCL Fig.6(b) Fig.6(c) Fig.6(d) 2 Fig.6(e) Konolige [4] k P k P = {p 1, p 2,... } (5) 8 10

19 Fig. 6 11

20 F (P ) ( ) F (P ) = i I(p i ) + i A(p i, p i+1 ) (6) I A I A p i p i+1 2 A Fig 8 ( )C = 20 I = 5 A = 39 Fig.10 Fig. 7 Waverront 12

21 Fig. 8 Fig. 9 Fig. 10 lpn 13

22 m p l t n m p (7) m p = 1 n ( ) n li i=0 Fig11(a) 8 Table1 Fig11(b) t i (7) Table

23 (a) (b) Fig. 11 : (a) (b) m 15

24 (8) m i = 1 ( ) n 1 n i=0 l 2 i (8) Fig. 12 (Fig.13)

25 Fig Fig.14 d

26 Fig. 14 )9) cost = 1 t t M i M i 1 (9) i=0 M ( 1 0) t

27 Carmen Carmen Tool Kit [5] Carmen A 2 ˆ ˆ ActivMEDIA ROBOTICS Pioneer3 Fig.15 SCIK LMS 200 Fig.3 Table. 2 Pioneer [m] x 0.42[m] 190mm 0.5 [m/s] 40[deg/s] 16.5[kg] 8 20 [deg] 4.0 [m] 19

28 Fig. 15 Pioneer3 Table (1σ) ms 10 mm 15 mm 5 mm 1 80 m 10 m 20

29 (A) (B) (C) (D) (E) (F) (G) (H) (I) (J) (K) (M) Fig. 16 実験環境:本研究で実験を行った環境地図上の記号と写真が対応している写真は矢印の方向へ見たものである 21

30 F Fig.17 CAD CAD ICP 22

31 23 Fig. 17 CAD ; CAD

32 (Fig.18) [m/s] 40[deg/s], 0.5[m]

33 25 Fig. 18

34 3.3.2 Table.4 0 Table. 4 [s] [m] [deg] [ ] Fig Fig19(b) Fig19(a) 4 26

35 + (a) (b) Fig. 19 (a) (b) Fig

36 Fig Fig

37 Fig. 23 Fig

38 Fig Fig.25 Table cm 10cmx10cm 10cm Table. 5 [cm] [cm] A B C (a)(f)

39 Fig. 25 ソナーマップ; 黒色で描かれているのがレーザから生成した地図赤色で示されているのがソナーから生成した地図青色は不明な領域を示すまた図中のアルファベットは Fig 中の物と対応している (a) (b) Fig. 26 ソナーマップ寸法評価 31

40 Fig

41 Table. 6 [ ] 10/10 0/10 (a) (b) Fig

3.5 1 (Fig.30) 1 Table.7 Fig.29 10 2 Table.

42 3.5 1 (Fig.30) 1 Table.7 Fig Table. 7 [ ] [s] 10/ / (a) (b) Fig

43 Fig

44 Fig.16 Fig.27 1 Fig.30 Fig.31 S A E ˆ 1 0.5[m] ˆ Fig

45 3 Table.8 12 Fig A 2 Fig16(K) E B Fig16(G) 1 (Fig.30) C Fig16(M) D Fig16(F) Fig16(E) A E B 37

46 Table. 8 A [m] [s] Fig. 32 A Table. 9 B [m] [s]

47 Fig. 33 B Table. 10 C [m] [s] Table. 11 D [m] [s]

48 Fig. 34 C Fig. 35 D 40

Table. 12 E [m] [s] 1 2 41.98 85.59 14.78 34.53 14.

49 Table. 12 E [m] [s] Fig. 36 E 41

50 [m] 291.5[s] 0.47[m/s] Fig

51 4. Pioneer2 43

52 ,,..,.,,.... OB 2,...,,.,,,,., 44

53 ,. 45

54 [1] P.J. Besl and N.D. McKay. A method for registration of 3-d shapes. IEEE Transactions on Pattern Analysis and Machine Intelligence, 14(2): , [2] PE Hart, NJ Nilsson, and B. Raphael. A Formal Basis for the Heuristic Determination of Minimum Cost Paths. Systems Science and Cybernetics, IEEE Transactions on, 4(2): , [3] Y. Hosoda, S. Egawa, J. Tamamoto, K. Yamamoto, R. Nakamura, and M. Togami. Development of Human-Symbiotic Robot EMIEW -Design Concept and System Construction. JOURNAL OF ROBOTICS AND MECHATRONICS, 18(2):195, [4] K. Konolige, SRI Int, and M. Park. A gradient method for realtime robot control. In Intelligent Robots and Systems, 2000.(IROS 2000). Proceedings IEEE/RSJ International Conference on, volume 1, [5] M. Montemerlo, N. Roy, and S. Thrun. Perspectives on standardization in mobile robot programming: the Carnegie Mellon Navigation (CARMEN) Toolkit. In Intelligent Robots and Systems, 2003.(IROS 2003). Proceedings IEEE/RSJ International Conference on, volume 3, [6] E. Rimon and DE Koditschek. Exact robot navigation using artificial potential functions. Robotics and Automation, IEEE Transactions on, 8(5): , [7] S. Thrun. Robotic Mapping: A Survey. Exploring Artificial Intelligence in the New Millennium, [8] S. Thrun, D. Fox, W. Burgard, and F. Dellaert. Robust Monte Carlo localization for mobile robots. Artificial Intelligence, 128(1-2):99 141,

55 [9] Sebastian Thrun, Wolfram Burgard, Dieter Fox, and ( ).., [10],,,,,,,, and. 1p2-n-036 i (, )., 2005:98, [11]. wakamaru (,[ ] )., 109(1051): , [12],,,, and.. In 2007, pages 2A1 D06(1) (4),, [13],,, and. (, / )., 49(3): , [14], editor. (2007 ). Number 34 in JILPT.,

56 A. CARMEN Toolkit A.1 Carmen CMU GPL ver0.74 A.2 A.2.1 Debian 3.1 sarge Debian 4.0 etch Ubuntu 7.10 Ubuntu 8.04 Ubuntu 8.10 gcc ver4.1 ver4.2 -Werror A.2.2 libgtk2.0-dev gdk-imlib11-dev libwrap0-dev libmagick++9-dev libmagic*-dev swig 48

57 python python-all python-all-dev python-tcpwrap python-cxx-dev A.2.3 ˆ /src/joystick/joyctrl.h 54 #define CAR- MEN JOYSTICK DEVICE /dev/js0 #define CARMEN JOYSTICK DEVICE /dev/input/js0 ˆ pioneer /src/base/pioneerlib/pioneer libċ A.3 ˆ central carmen ˆ param deamon ( ) ˆ simulator laser,pioneer ˆ laser ˆ pioneer pioneer ˆ localize ˆ navigator ˆ robot 49

58 Fig. 38 carmen 50

59 A.4 A.4.1 $ central $ param daemon -r ROBOTNAME MAPFILE $ simulator $ localize $ navigator $ navigatorgui $ robot $ robotgui Fig. 39 navigatorgui A.4.2 $ central $ param daemon -r ROBOTNAME MAPFILE $ laser $ BASEDRIVER( ex. pioneer) $ localize $ navigator 51

60 Fig. 40 robotgui $ navigatorgui $ robot $ robotgui PC A.5 $ logger file name.clf * $ log2log $ vasac 52

soturon.dvi

soturon.dvi 12 Exploration Method of Various Routes with Genetic Algorithm 1010369 2001 2 5 ( Genetic Algorithm: GA ) GA 2 3 Dijkstra Dijkstra i Abstract Exploration Method of Various Routes with Genetic Algorithm