ª Œª Œ 26ƒ 1D Á 2006 1œ pp. 195~202 ˆ «~ ª xœ k» w en Analysis of Infiltration Route using Optimal Path Finding Methods and Geospatial Information ûáx Á y³á Bang, Soo NamÁHeo, JoonÁSohn, Hong GyooÁLee, Yong Woong Abstract The infiltration route analysis is a military application using geospatial information technology. The result of the analysis would present vulnerable routes for potential enemy infiltration. In order to find the susceptible routes, optimal path search algorithms (Dijkstra s and A*) were used to minimize the cost function, summation of detection probability. The cost function was produced by capability of TODG (Thermal Observation Device), results of viewshed analysis using DEMG (Digital Elevation Model) and two related geospatial information coveragesg(obstacle and vegetation) extracted from VITDG(Vector product Interim Terrain Data). With respect to 50m by 50m cells, the individual cost was computed and recorded, and then the optimal infiltration routes was found while minimizing summation of the costs on the routes. The proposed algorithm was experimented in Daejeon region in South Korea. The test results show that Dijkstra s and A* algorithms do not present significant differences, but A* algorithm shows a better efficiency. This application can be used for both infiltration and surveillance. Using simulation of moving TOD, the most vulnerable routes can be detected for infiltration purpose. On the other hands, it can be inversely used for selection of the best locations of TOD. This is an example of powerful geospatial solution for military application. Keywords : geospatial information, optimal path, infiltration route analysis, Dijkstra, A* en xœ» y w wù. en w w. ƒ w en» w k y w tx w yw š ( p A*) w., eš x w ƒ x (VITD) sw xœ f (coverage) 2 f w w w. 50m 50m (cell) j» ƒƒ š, w yw ü. w xw. x p A* š j ƒ, A* š w d w. w en ƒ d y. e å ww, ƒ w en ý. d e kw» w. w w xœ y w w ƒ ƒ. w : xœ,, en, p, k 1. xœ x ƒ š. 2000 d l 1m ü šw z ƒ w, y(ÿw,, Ÿ ) w w t p w ƒ w. xœ w xš y w w w» z x š ƒ ww ƒ y y (Meng, 1988; John, 2001; Helgason, 2001). œ w Ÿw ƒ t ùkù w p t w, w zy lœw (E-mail: snbang@add.re.kr) z w zy lœw (E-mail: jheo@yonsei.ac.kr) z w zy lœw (E-mail: sohn1@yonsei.ac.kr) w (E-mail: lpllyw@add.re.kr) 26ƒ 1D 2006 1œ 195
m, w, w w. xœ y t y w en, ùw,, ƒ y (, 1994; w, 2000). ƒ x œ (NGA; National Geospatial-intelligence Agency) t t wù x (VITD; Vector product Interim Terrain Data) sw x eš x w en w» w wš x mw y ƒ. z(path planning) œw,» z, w z y š. Meng (1988) l wš y w wš, 3 ù yƒ v w A* p š wš w (Voronoi diagram) w. Helgason (2001) y w w, John (2001) w z yw» w ƒx wš v (integer programming) (genetic) š w. Niederberger (2004) y ƒx txwš A* š w ü š / w š w. ƒx w t wš A* š (greedy method) yww w,» A* š w š., l ƒx w š j» w j ú ƒ j. w ù» y w w x w p ƒx y (grouping)w w y w» w š., Kwok (1999) š w x» w» /» ƒ 2 l y w v (dynammic programming)» w, Howard (2002) x p 5 l r (fuzzy map) wš A* š w w k»(planetary rover) w. Marti (1994) šw x p w w» w ArcGIS vp w l z sƒ w w, ù x p w ww. ƒ e y en xk l y w w. w ful y ƒš», v w A* š x š 196 w w ü ww. (TOD; Thermal Observation Device)ƒ e y xœ sw ù x w šw xp w lxk y wš» w k y ƒ en w w. l k y w» w vp (ArcGIS Software)» y w, y š w v Visual C++ w x ww. w t A* š (Hart, 1972) v p š (Dijkstra, 1959) w w. 2. 2.1 en w y. xk e y ƒ w xp š w en w t w. x w ƒ wš en w 1. 1 ew s w ew s w. k y w Fƒ en w t w. 2 w š. x VITD, 1:5,000 e š w 1. en (ƒ ; ) ª Œª Œ
2. eš x w. w 2 s œ j»(50m 50m) ƒx w wš, ƒ w en k y l k y w. (Thermal Observation Device) TAS-970K w k y ACQUIRE w w. ACQUIRE d wƒ (MRTD; Minimum Resolvable Temperature Difference) w,» p t k y w (y, 2003). ú x w. 2.2 VITD NGA ³ w ³ l t y w» w x w wš t xœ l q x. 3 (OBS; Obstacle), / x (SLP; Slope/Surface Configuration), m /m (SMC; Soil/ Surface Materials), (SDR; Surface Drainage), (TRN; Transportation), (VEG; Vegetation) 6 f (coverage), ƒ f. OBS f, w, y ( ),, œ x sww. SLP f txw x sww. % tx, x j» 5,000s l(50mü100m). SMC f m Ÿ, m w m txw. SDR f, w,, y x w txw. TRN f,,, l y m( ) x txw. VEG f m m ü ƒ txw. OBS f VEG f sw w. OBS f ƒ w w, VEG f ù s w. TRN, SLP, SMC» sw ƒ q w y w, SDR f (w, 2000). w d e ƒ e. w w t tš ù ù tx DSM (Digital Surface Model) w ù, x yw DSM w y w». VITD (œ w 50m 100m) š w 3. x (VITD) 6ƒ f 26ƒ 1D 2006 1œ 197
1:5,000 e š w 50m eš x w ( 5). 3. k y k y 2 w sy, ƒ, ƒ TOD k y w w. sy w x s w en y t w, ƒ ü w ƒ w w. ƒ TOD ƒ e w x w s w, TOD k y k y w. w 50m 50m wù l y w. ƒ w en ƒ y š w TOD k y w. ƒ x w w,» k WGS84, n t UTM 52 zone yw w. y t vp qj ArcGIS 9.0 w w. 3.1 sy sy VEG f sw 3, VEGAREA(vegetation area), VGFAREA(vegitation forested area), VGWAREA(vegitation water area)» w. VGFAREA sw DMT(Density Measure of % Tree/Canopy Cover) w en y w. DMT w s y txw t 1 t 25%, 50%, 75% 4 g. x ƒ sy w. VEGAREA VGWAREA s w y ù š» DMT ƒ û w w. 4 sy û, sy ùkü. 3.2 ƒ TOD k y TOD e x š w k y w» t 1. sy Layer DMT sy VEGAREA - 0.125 VGFAREA 0~25% 0.125 VGFAREA 25~50% 0.375 VGFAREA 50~75% 0.625 VGFAREA 75~100% 0.875 VGWAREA - 0.125 w, TOD k y wš eš x l ƒ ww w w. 5 x TOD e ùküš, 6 TOD e e l k y. TOD e e 4 k y txw, ƒ k y û w. 7 ƒ TOD e ƒ w š. ƒ x s w» w 2.2 w eš x w, ArcGIS ƒ (viewshed)» w. 7 t TOD d ƒ w, d ƒ w. 6 TOD k y 7 ƒ w w x š w TOD k y ƒ e w ( 8). TOD ƒƒ w k y w k y š w y wœ (1) w 4 k y wù k y mww ( 9). P( A B) = P A 4. sy 5. TOD e ( ) + P( B) P( A B) (1) 198 ª Œª Œ
6. TOD k y» P(A) P(B) TOD w k y w, P(A B) TOD k y, P(A B) TOD wù TOD k y w. 4 k y mww» w 2 TOD w k y mwwš ù 2 TOD w k y mww, mw 2 k y mww k y w. 9 t k y w 4 TOD wù TOD w en ƒ k y w. 3.3 k y k y w» w 3.1 w sy C(x, y) 3.2 w TOD k y D(x, y) w w.» x, y 2 e txw. TOD k y D(x, y) k y ƒ, sy C(x, y) k y. sƒ x k y 1.0, 100% sƒ k y 0 š w k y (2) txw. y w w» w (3) w. k y 10. P x, y ( ) ( 1 C( x, y) ) P x, y 9. TOD k y ( ) = D( x, y) ( 1 C( x, y) ) (2) (3) 3.4 p x OBS f OBSLINE(obstacle line) 7. TOD e ƒ 8. ƒ TOD k y w 26ƒ 1D 2006 1œ 199
10. k y F_CODE(feature code) OBSAREA(obstacle area) w ƒ ƒ w. t 2 F_CODE ƒ / ƒ š, 11 w ƒ w. p s w w» w, y. VITD ³ e(dragon Teeth), w (Depression), t (Geographic Information Point), (Wall), q v(pipeline/pipe), e (Location Category), w (Moat), / / (Bluff/Cliff/Escarpment), (Cut), (Embankment), y (Volcanic Dike), k (Hedgerow), (Ramp), t 2. p Layer F_CODE p OBSLINE AL260 (Wall) ƒ ƒ OBSLINE DB010 / / (Bluff/Cliff/Escarpment) OBSLINE DB070 (Cut) ƒ OBSLINE DB090 (Embankment) ƒ OBSAREA DB080 w (Depression) ƒ sw w p w. 4. w k w kw. x l 2 ƒ w 8 w v š ƒ w. x 500 500 = 250,000 y ù œ w ƒ w f. ¾ k (depth first search)ù s k (breadth first search) w š j ƒw,» (greedy method) wù w» w (Horowitz, 1978). k w» w ƒ š p (Dijkstra, 1959) š w ü z A* š (Hart, 1972) w xwš w w. A* š w f = g + h w.» g ù w w, h x ¾ w x w. p š w. Russell (1995) A* š x w ƒ sƒ z w. A* š,, w sƒ x w ƒ ƒ. w g (4) ù k y w, w h (5) x» k y μ detect txw. μ detect w w. g = ( x, y) A P( x, y)» A ù w w. (4) h = y Y goal μ detect (5) 11. ƒ 200» Y goal 1 w t w y t w, y Y goal t ¾ û w. t w ƒ w w. 5. x w 7 w k w. 12 p š w w, 13 μ detect =0.01 A* š. t 3 ùkù 1 w en ƒ 4.60, 4.63 š ª Œª Œ
t 3. en k y y p A* 1 4.60 4.63 2 4.69 4.71 3 5.62 5.64 4 6.23 6.25 5 7.16 7.18 6 7.75 7.76 7 7.45 7.47 12. p š. k y f w f ƒ. t 4 š w w k y w. A* š p š w w., ³ š p š w w, j y š œ w ƒw w w š w A* š w. t 4. w k y μ detect ( : ) k y A* (0.01) 8.27 4.63 A* (0.02) 3.17 5.06 A* (0.03) 1.77 5.91 13. A* š (=0.01) ƒ û k y. A* š p š w. μ detect y g A* š 14. μ detect f t w w w., û ƒ w e w f û w ww w, k y f A* (0.10) 1.22 6.76 6. Dijkstra 15.87 4.60 en œ en w y ƒ w, e w y w. mw xœ w l y wš en y w 14. k 26ƒ 1D 2006 1œ 201
w, x mw y ƒ. w p š A* š w fu l w en z. yw en w y w yw ƒ v w, x en» w ƒw y w en» g w. w y w q w y ƒ w en ƒ w w. š x, x, y (1994)» w, w xœ wz, w xœ wz, 2«, 1y, pp. 107-119. w, ½ y, (2000) p, š, KTRC-409-000221L, w y, ½,,, x(2003), xsƒ š, TEDC-317-030544, w Dijkstra, E.W. (1959) A note on two problems in connection with graphs. Numeriche Mathematik, Vol. 1, pp. 269-271. Hart, P.E., Nilsson N.J. and Raphael, B. (1972) A formal basis for the heuristic determination of minimum cost paths. SIGART Newslett, Vol. 37, pp. 28-9. Helgason, R.V., Kennington, J.L. and Lewis, K.R. (2001) Cruise Missile Mission Planning: A Heuristic Algorithm for Automatic Path Planning, Journal of Heuristics, Kluwer Academic Publishers, Vol. 7, No. 5, pp. 473-494. Horowitz, E. and Sahni, S. (1978) Fundamentals of Computer Algorithms, Computer Science Press, Maryland, USA. Howard, A., Seraji, H. and Werger, B. (2002) Fuzzy terrain-based path planning for planetary rovers, Fuzzy Systems, 2002, FUZZ-IEEE'20, Proceedings of the 2002 IEEE International Conference on, IEEE, Vol. 1, pp. 316-320. John, M., Panton, D. and WhiteS K. (2001) Mission Planning for Regional Surveillance, Annals of operations research, Baltzer, Vol. 108, No. 1/4, pp. 157-173. Kwok, K.S. and Driessen, B.J. (1999) Path planning for complex terrain navigation via dynamic programming, American Control Conference, 1999, Proceedings of the 1999, IEEE, Vol. 4, pp. 2941-2944. Marti, J. and Bunn, C. (1994) Automated Path Planning for Simulation, AI, Simulation, and Planning in High Autonomy Systems, 1994, Distributed Interactive Simulation Environments., Proceedings of the Fifth Annual Conference on, IEEE Comput. Soc. Press, pp. 122-128. Meng, A.C.C. (1988) AMPES: adaptive mission planning expert system for air mission tasks, Aerospace and Electronics Conference 1988 NAECON, Proceedings of the IEEE 1988 National, IEEE, pp.1225-1231. Niederberger, C., Radovic, D. and Gross, M. (2004) Generic path planning for real-time applications, Computer Graphics International 2004 Proceedings, IEEE, pp. 299-306. RussellS SU and NorvigS P. (1995) Artificial intelligence: a modern approach. New Jersey: Prentice-Hall. http://www.nga.mil/staticfiles/ocr/nga_history.pdf, NGA History, National Geospatial- Intelligence Agency, St. Louis, (last date accessed: 16 December 2005) ( : 2005.10.7/ : 2005.12.1/ : 2005.12.16) 202 ª Œª Œ