NASA Technical Memorandum A Workstation Based Evaluation of a

NASA Technical Memorandum 102882 A Workstation-Based Evaluation of a Far-Field Route Planner for Helicopters David N. Warner, Jr. and Francis J. Moran (NASA-TM-I02882) A WORKSTATION-BASED EVALUATION FAR-FIELD ROUTE PLANNER FOR HELICOPTERS (NASA) 24 p N92-33609 OF A Unclas G3]04 0117766 June 1991 National Aeronautics and Space Administration NASATechnicalMemorandum 102882 A Workstation-Based Evaluation of a Far-Field Route Planner for Helicopters David N. Warner, Jr. and Francis J. Moran, Ames Research Center, Moffett Field, California June 1991 National Aeronautics and Space Administration Ames Research Center Moffett Field, California 94035-1000 COLOR ILLU 'f SUMMARY Helicopter pilot. of automated algorithm planner ments. dynamic flight route missions planners. at very low, Nap of the Earth, Ames (far-field) Research route preflight altitudes mission threats place a heavy workload various planner on the types To aid in reducing aids in selecting this workload, the overall Center has been During investigating planner, or change algorithm, a route the mission, As part of an automated a new route an evaluation the mission. from a flight to be flown. the route requireplanon the distance can be used to replan This report describes programming in case of unexpected of a candidate meets In general, time schedule, most in mission based for route route planning the requirements techniques. and during This algorithm of the requirements within ning, both preflight are to minimize and/or fuel and the deviation of available fuel and time. and must be flyable the constraints INTRODUCTION Helicopter workload about Nap pilot, effort flights are frequently One flown in areas and under hazardous mission level. regime flown The ground conditions where a heavy from as is known is placed on the pilot. treetop or NOE. example is a military at very low altitudes, latter elevation automation 50 m above of the Earth, both at NASA, level to about Successful planning Center 2 - 6 m above operations for several in this flight years (refs. requires to aid the a major of in preflight Ames mission Research and in the cockpit. This area of automation has been 1-6). Part of this effort is the study automated route planners, which fit into three categories: far-field, field planners determine the overall route to be flown, e.g., which knowledge ners, through of the terrain the widely (refs. planner planner the terrain based spaced on a coarse, waypoints system, digital terrain database utilizing and closely which which relies spaced mid-field, and near-field. Farvalley to fly through, by having (refs. 3,4). Mid-field database route planterrain data, plan a safe route are dealt with by (ref. 6). It is typi- digital 3,5). Obstacles do not appear in the digital sensors a near-field A mission cally known goals, during or guidance on on-board for its inputs planning. is used by the mission tool which weather goals. route of the destination disturbances, planner. describes commander location, for pre-flight any intermediate stores, of this mission mission waypoints planner an interactive man-made route computer locations threats, mission allows the user to view weapons planner a graphical map of the operational or destinations, time constraints is computation inappropriate route or of as new planner area and to input fuel capacity, and any other a flight mission Part of the operation A route a new route the "best" information algorithm. by a far-field of this type could one becomes evaluation also be used by a pilot to compute This report if the preplanned a qualitative is received. of a candidate The far-field of a Small route planner Innovative described Research in this report (SBIR) was provided As designed to Ames Research Center as part Business contract. and implemented by the contractor on a personal computer of a full mission planner. Although (ref. 5), a graphical user interface gave the system many features parts of the graphical user interface were implemented at Ames on a SUN workstationalongwith the routeplannersoftware,theprimaryemphasisat Ameshasbeen to useandevaluatethe routeplanningalgorithmindependent f otherhigher-levelmissionplanning o aspects. his reportdescribes routeplannerin generalterms,because proprietaryrestrictions, T this of andits performanceastestedon a SUN workstation.Also includedis a brief descriptionof the graphicaluserinterfacewhich wasusedon the workstationto define inputsto the planner,execute the routeplanneralgorithm,anddisplaythe resultingroute. FEATURES DESIRED IN A ROUTE PLANNER A route planner has one overall function to perform, that is to compute an acceptable route from a designated starting location to one or more optional, intermediate nated final destination. The criteria which def'mes the acceptability text of the route sions, a planned planning route and some criteria which must ensure that a helicopter will arrive goal locations and to a desigof the route depends on the conhelicopter NOE miswith a high proba- must be met. For military at its destination bility of survival. In general, from a flight time schedule, Some dezvous missions include with other aircraft an acceptable route minimizes the distance and/or fuel and deviation and must be flyable within the constraints of available fuel and time. to meet time goals location. computer at specified for route locations, should replanning for example during the to renor to pass a control on an airborne Also, the algorithm run fast enough the requirement (within about mission. one minute) to be used In order to accomplish and known threats, either far-field relatively planner coarse these requirements, a far-field planner must have knowledge of the terrain weather to be avoided or man-made threats such as missile sites. Since a a general route to be flown, the terrain can be defined depending by a on the of 0.25 to 2 km between data points only determines grid of data, on the order terrain's topology. Threats can be made known to the planner by specifying their locations and a model which defines their range and other pertinent parameters. From the model and the terrain topology of survival. the inputs at the threat Some location, terrain must masking of the threat locations can be used which to maximize are along which the probability as well as Once the criteria the route. meets capability locations be made for user input of starting or targets and ending locations any intermediate for acceptability. such as rendezvous algorithm are known, a computerized can be used to plan a route GRAPHICAL USER INTERFACE Inputs provided a SUN to and control by the graphical Along of the route user planner algorithm and display of the resuiting is a special to select planned purpose various route window func- are on interface, figure 1. This user interface subwindow workstation. the top of the upper are buttonsused tions. From left to right, the top row of buttpnsstart able vs distance form, select various map options, added features from the map, select prestored the program. placement The icons on the second the route planner algorithm, plotresults in varicontrol display of a grid on the map, clear usercontrol types from a route weighting points function, or threats starting and quit for location, of control scenarios, represent, 2 row are used to select on the map by the user. The icons left to right, the route a non-route-optimized routecontrolpoint, an optimizedroutecontrolpoint, a target,the destination, an adverse weatherregion,a missilesite,anantiaircraftgunlocation,anda radarsite.Routecontrol pointsasusedin this reportarelocationswhich the userdesires thatthe routepassover.Theselocations may be rendezvous oints,navigationupdatelocations,passenger p dropoff sites,or may be specifiedto force theroutethroughareas plannerwould not ordinarily select.The selectionbutthe tons andiconsareusedby moving the arrow-shaped cursorontothe desiredbuttonor icon by moving a mouseandpressing("clicking") the left mousebutton.The userthenmovesthe cursorto place the selected icon on the coloredmapin a lower subwindow(described shortly) andagainclicking the left mousebutton.Whenthis is done,the icon is displayedon the map,andits location is savedfor useby the routeplanneralgorithm. The second subwindowis whereroutelengthandnumberof waypointsalongthe computedroute aredisplayed.To the right is a smallsubwindowwheretherouteplanneralgorithm is identifiedand mapidentifier andscalingaredisplayed.Also shownarethe currentlocationof the cursorwhenit is movedin the mapsubwindow.At thebottomof the subwindowis the currentvalueof the route weighting factor,to bedescribedin a later section. The operationalareaof interestis displayedin the largegraphicssubwindow.Terrainelevations on the map arecolor codedasshownin the smallersubwindowbelowthe map,whereblue is sea level andthe othercolorsrepresentncreasing i altitudesegments from lowest(light green)to highest (darkmaroon)in the area.Routecontrolpointsandthreaticonsaredisplayedat locationsselected by the user.If desiredthe usercanclick on the "GRID" buttonto displaygrid lines on the map. To plan a route,the userfirst selectsthe propermap.Thenthe startinglocation,intermediate routecontrol points,target(s)location,andthe final destinationarepositionedon the map.Locations of anyknown threatsarepositionedon the mapandredlines areautomaticallydrawnto showthe threatregioncovered.Thenthe userclicks onthe "ROUTE PLANNER" buttonto executethe route planner.The routeplanneralgorithmis executed andthe computedroutedisplayedas a whiteline. Plots of terrainaltitude,cost,andcumulativecostalongtheroutecanbe displayedby clicking on the "PLOT" button. Prestored scenarios canbe selected clicking on the "FILE" button. by ROUTE PLANNER DESCRIPTION AND EVALUATION General The rithm route planner, called Dynaplan Description is based definition on a dynamic standard and masking, programming implementations and manipulations these and not algo- by the developer, threat (ref. 3). The implementation such as terrain by the contractor data handling, departs from more only in a few areas, of the cost data to reduce features will be described also describes used some in the Ames system. computation speed. Because of the proprietary nature of the software, below only in general terms. Reference 1 describes the route planner planning aspects of the contractor's implementation which were mission The generalprocedures usedin implementingthe dynamicprogrammingalgorithmto computea near-optimumrouteis asfollows. First, anarraywhich definesthe flight environmentis initialized to the terrainelevation.Thenthe threatmodelsselected andlocatedby theuseraresuperimposed on the terrainarray.The three-dimensional transitioncostarrayis generated eightcontroldirections for ateachcell by computingthecell-to-cell transitiontimes.Thesetransitiontimes areapproximations sincecell-to-celldistanceis an approximationandthegroundspeed assumed is constant.After the statespaceparameters defining the operationalsituationarethus defined,the dynamicprogramming algorithmis executed computethenear-optimumroute.A loop structureis usedto computethe to route asseparate near-optimum segments from the startinglocationthroughthe intermediate controlpointsandtargets,if any,to the final destination.Someof thesecomputationalstepswill be describedmorebelow asresultsof theevaluationis described andillustrated. Terrain Data To represent the terrain in the dynamic programming grid, digital terrain elevation data (DTED) used for this study was obtained from the U. S. Geological area in northern California, called here the Napa data. The seconds, planner which algorithm, is approximately an area were used. 100 m north-south data points in figure planning, 80 x 80 km was selected, The elevation route as shown Survey. spacing It is a one degree by one degree between data points is three arcFor testing 1201 of the route x 1201 elevato ( 167 m) and 80 m east-west. so a subset 1. Since of the original data tion data points is not deemed for this area were reduced terrain the data was further by interpolation 480 x 480 data points necessary for display, at this resolution compressed for far-field for test pur- poses to 40 x 40 (2 km spacing), 80 x 80 (1 kin), 160 x 160 (0.5 kin), and 240 x 240 (0.33 kin). The altitude was set to 75% of the maximum to minimum elevations in each subarea, or cell, figures 2-5 based on empirical evaluations. programming requirements. ceils are initialized or in some The contractor to the compressed form of scaled units the elevations terrain elevations. These valmemory to The dynamic and disk storage ues may be elevation in feet or meters to minimize computer normalized to 0 to 255 to correspond one byte to minimize computer memory requirements and to reduce Normalized terrain data was used for these tests as well as elevations effects on the algorithm. Threats program execution time. in meters to evaluate any Threats implemented and a radar site. Generic for instance, values defined plished were figure whose magnitude in the test system include adverse weather, a missile, an antiaircraft models for each threat were used since selection of different missile to evaluate value the route planner algorithm. The threat models is a relative of the threat lethality and which decreases linearly gun, types, to the was not required are simply ranges. Ranges when the threat initialized 6. areas used are listed below in table 1. Threat masking by the terrain is accommodel is applied to the appropriate dynamic programming cost cells which elevations. threats. A separate threat intervisibility displays algorithna was used to disin Examples of the threat and their effects are shown to the terrain around play the lethal 4 Table 1.Threatranges Threat Anti-aircraft gun Missile Thunderstorm Range(kin) Transition The cell-to-cell change function elevation and distance are possible. change transition cost computed Costs algorithm is a function as the product of elevation the cost of the average by the planner from the current As delivered, cell to the adjacent the transition distance, C t = F_, * D t t cell. Various ways of implementing cost was computed and the transition where C t is the ceil-to-cell as previously distances. transition from mentioned, cost, center E t is the average to center elevations elevation were of the current To reduce cell transition and adjacent tar- get cells, Another ied from and D t is the distance simplification of the two cells. and east-west computer distances distances more memory and 141 which vareleva- requirements as the diagonal terrain normalized to the range of 0 to 255. was to use 100 as north-south For this test, the terrain terrain favors elevations lower This cost function data had actual penalizes cell transition changes 0.33 to 2 km and actual of 5.6 meters. that the algorithm from 0 to 1427 m. This gave elevation whenever routes possible; an equivalent i.e., it is a tion quantization with the result than distance elevation valley-seeking algorithm. An example of the algorithm's operation using normalized, 1 km spaced terrain data is shown in figure 7. The route chosen followed the lowest elevation possible, a distance of about 117 km. In order parameters to assess the effects of the elevation measures and distance of elevation approximations, and distance identical confirmed the data used for the the new in figdiffer- was changed to be correct in meters. Using numbers gave the result shown in figure 8. The route appears ure 7. Examination of numerical data from the computations ences Because were present. This result confu'ms that some of a lack of adequate situations, by visually location. valley, time, an exhaustive like that shown required. the terrain 9 shows in a distance verification in figure to the first one shown that only very minor approximations in cost computations are valid. was not done however. 7 is undesirable if a shorter the algorithm's to force route is at In many available computation some through a long route altitude examining Figure resulting with only small changes The pilot can influence and placing control-point 36 km. route the route display an intermediate placed control-point appropriate a different an intermediate of only about 5 Although the pilot shouldhavethe capabilityto placecontrol-pointsat arbitrarylocationsas described above,a moregeneralmethodof weightingthe routecalculationbetweenminimizing distanceandminimizing altitudeis desired.Onerathersimplemethodto dothis wastestedusing a weighting factorWEDto modify thecell-to-cell transitioncostto Ct = {(10 -WED). Et} + (WED* Dt) / where the weighting distance. factor Figure WED is adjustable 10 shows route were three (33.74 by the pilot of varying 10 from 0 for minimum WED. The longest the undesirable short valley elevation route effect (about to 10 for kin) is of going 34 km). that the eleimplementa- minimum over Only vation results (114.45 for WED = 0. The shortest the highest minor versus terrain differences distance km), WED -- 10, produced from 2 to 8 all gave of weighting. some in the area. Values noted give more control routes for values of 2, 4, 6, and 8, indicating Nevertheless, user control that a better the test proved of route tion of the equation might weighting concept can provide characteristics. Effects To investigate nario representing were included man-made effects. the effects of varying situation route of Terrain terrain Data Spacing on the route was used, through location planner figure algorithm, a sceby to the data spacing route an operational The computed with a defined target 11 (a). No threats unhindered the low valleys to allow the algorithm to compute goes from segments the terrain through the starting target and then to the final destination. The first segment is from the starting destination. For the first test case, 0.33 shows used. km, figure this route 1 l(a). drawn shows the route As described earlier, the route is computed in two segments. location to the target; the second, from the target to the final was computed displayed are almost following direction. using terrain data in a 240 x 240 grid spaced distance is 54.5 km. Figure in figure 11 (a). Visual at The map display on the terrain it generally map and route is 480 x 480. The route at the same identical the lowest Between through 1 l(b) algorithm inspeclocaThis 240 x 240 grid as the planner as expected, This displayed to that shown terrain, the target tion of the route tion to the target takes a fairly from the starting two higher areas. in a southeasterly and the final destination, the route to the north. distance than direct path over a low rise and down a pass between route segment chosen was unexpected, since a lower route is available a short distance The chosen route had a lower cost due to the small terrain elevation change for a short the somewhat was longer, but lower altitude, route operating terrain data would have had. Elapsed compute 125 seconds, Using including the UNIX system gives (SunOS) the results overhead. in figure time for this test 0.5 km (160 x 160) spacing i21 Comparing fig- ure 12(a) to figure 11 (a) shows that the flu'st route segment is offset a small distance from the lowest terrain elevation. The route distance is 54.3 km. In figure 12(b), the route still appears to be very close to the lowest When the terrain terrain data elevation. Elapsed compute time was 50 seconds. to 1.0 km (80 x 80), the route offset becomes spacing was further increased more pronounced when viewed on a high resolution display, as shown in figure 13(a). The route distance is 54.0 kin. When viewed at the same resolution as the planner algorithm used, the route 6 appears be still roughly centered the lowestpart of the terrain,asshownin figure 13(b).Elapsed to in computetime was 10seconds. Theseroute offsets algorithm somewhat, small vation. or weighting but some elevation. data are consequences factor offset for reducing is inevitable, Operationally, among could, persons depending using of less accurate the terrain since this offset Lower planner data would level a relatively representation large likely route move of the terrain. probably A different change to determine using higheleif however the offset a as is the grid size would area is processed be of minor guidance the route systems closer cell' s weighted resolution they view terrain Pilots importance, distancedifferences or other the routes. the route image. on their algorithms, to the lowest may find the offset objectionable, the map as a high resolution Another consequence of increasing rain. As closely spaced data is averaged for use by helicopters help minimize probability of detection. the terrain to reduce data spacing is the reduced knowledge of the terthe grid size, knowledge of narrow valleys suitable planner's capability fo r finding routes to is lost. This loss reduces the route SUMMARY OF RESULTS The far-field planner targets whenever around alternate some long when route planner which has been described threats has most of the features routes to user specified possible to avoid appear appear changes which whenever desired in a route and to minimize of this type. It computes generally The routes will allow however, acceptable avoid moderate routes control lower locations altitudes and to a final destination. programmed possible high regions. shorter cost function but generally Occasionally routes indicates and appear the cost criteria into the algorithm, i.e., the routes altitude are computed to follow long excursions to be excessively testing could of an provide with moderate altitude increases are available. The cursory that a different implementation of the cost function improvement. algorithm appears to make good been and route planner. lacks some of the requirements a helicopter would planner. to meet best be performed This route which distance, is in assessing, by a higher planner is horizontal or providing level mission comuse of the terrain The question answered requirements. elevation about what data and to efficiently resolution the planner of terrain on algorithm which since it is so dependent The planner integrate elevation the terrain computes should knowledge of threats into the cost matrix. data is required characteristics a route has not necessarily and the mission by this study, However in 125 seconds for a far-field the tested a route's a route need or less for resolutions route of 0.33 to 1 km terrain data spacing, be acceptable One area where information Perhaps planner, putes testing algorithm capability meets to assess, whether to allow time and fuel constraints. algorithm constraints functions planner but it would information as linear for altitude meets from the route changes. flight time and fuel usage route of route defined distance to assess only and does not account whether any specific ied in this evaluation. This simplification may not be adequate mission time or fuel constraints, but was not stud- 7 REFERENCES . Swenson, H. N.; Paulk, C. H., Jr.; Kilmer, Design R. L.; and Kilmer, Terrain-Following Systems F. G.: Simulation Helicopter Meeting, Evaluation Operations. Springs, of Display/FLIR Concepts for Low-Altitude, AIAA/AHS/ASEE Aircraft CO, October 14-16, 1985. Clement, W. F.; McRuer, CR-177453, and Operations Colorado . D. T.; and Magdaleno, (NOE) 1987. February R. E.: Some Data Processing of Rotorcraft. Requirements for Precision NASA Denton, Nap-of-the-Earth Guidance and Control , R. V.; Pecldesma, Research N. J.; and Smith, for Helicopters. F. W.: Autonomous NASA CR-177478, Flight August Mission and Remote 1987. Planning Site Land- ing Guidance Berger, . K. M.; Abramson, NASA M. R.; and Deutsch, August 1990. O. L.: Far-Field For Heli- copters. CR- 177560, Guidance December . Pekelsma, N. J.: Optimal NASA CR- 177515, with Obstacle 1988. Avoidance for Nap-of-the-Earth Flight. P Cheng, V. H. L.: Concept Development of Automatic Guidance Avoidance. IEEE Transactions on Robotics and Automation, Larson, R. E.; and Casti, New York. J. L.: Principles of Dynamic for Rotorcraft Obstacle vol. 6, no. 2, April 1990. Marcel Dekker, Inc., . Programming. 8 ORIGINAL COLOR PAGE PHOTOGRAPH - [] [] -',in path = km = 182.71 in path = km = !89.12 in path = km = 189.12 ir, path = 182 91 97 97 PLANNER REGION PLANNER DI SPLAY x: h: 75198 745 m :TAU :NAPA : 68 :488 m, 453 Number of points Route distance, Number o{ points Route distance, Number of points Route distance, Number of points ,/: 16932 m, 182 hd_._t 5 : 65 76 75 18 15 28 25 30 35 4_ 45 56 55 68 88 :i:i:i:!:i; !:!:,.:::':: i:i:i:i:i:i i!iiiiii!il :+:,:,:,: i:!:}:i:i:i ii!iiiiiiii !iiiiiii!i!i! ,:,:,:,:,:,:,:, Figure l. Graphical user interface for route planner. 9 t_lt_J_'_,'-,_. PAGE COLOR PHOI-OGRAPH [F'.OLITE PLANNER] ] _ [ PL'-tT [f4F:I['] _ [ =-ILE J :N_PA :80 _PLAY:4e :-_: 79666 m, 478 ?2833 m, 437 131 m Figure 2. Graphical representation of 40 x 40 (2.0 km) grid terrain as used by route planner. 11 PF_ECED!NG PAGE" BLANK I_OT F;LMED ORIGfN,,:-L PAGE .COLOR PHOTOGRAPH :::::::;:::;: ii!iii :ON :NAP_ :80 ;PLAY:88 :79666 m, 478 :72833 m, 437 :131 m ;Oj i :,:,:@:,: iiiiiiiiiii!iiiiiiiii: ,:.:.:,:,:,:,:,>:,:. !!iii!!i!i!i!i!?iiii iiiiiii!!iiiiii!iiii iiiiii!!ii!!i!i!!iii i_ii!!i_i: Figure 3. Graphical representation of 80 x 80 (l.O km) grid terrain as used by route planner. 13 PRECEDING PAGE BLAI'JK NOi FILMED !iiii 18 15 28 25 :38 :];5 46 i 45 i 50 55 _ 50 65 78 75 8rj -80 iili -75 -78 -65 -60 % %5 T Figure 4. Graphical representation of 160 x 160 (0.5 km) grid terrain as used by route planner. 15 PRECEDING PP_GE BLAf-iE NOT FILMED ORIGINAL COLOR PAGE PHOTOGRAPH _ii!!ii _ION :NAPA "_-R: 80 ',PLAY :248 79666 m, 476 '/:72833 n',, 37 4 : 131 rn i:i:i: g(q :i:!:! :::::: 18 15 28 25 I,. 38 35 I --I".0 45 I 58 J 60 55 78 75 80 -80 -75 -78 -65 _::::: C-:-;. -6o -55 iiii!i :::::: -30 I::::: ..... . . _ _.: -. m ...... ..... .:,:.:,:,:, ,:,:.:.:,:, .:.:.:,:.:, Figure 5. Graphical representation of 240 x 240 (0.33 km) grid terrain as used by route planner, 17 PRECEI::_NG P_GE B,__f,,K_-'.O'i' F_,LMED ORIGINAL.P ,.:n"" A,., COLORPHOTOGR,_PH Figure 6. Threat examples and effects. (a) Route without threats, (b) route with threats. Ir T 19 PRECEDING F_AGE BLA;'_X [_Cq" F.I.,'_,'D ° ORIGINAL PAGE COLOR PHOTOGRAPH • :.:':. Route dist._r, ce, N,_,ber of points Route distance, N,_Jber oq points points _m = 116.99 in path = 102 _m = :35.97 in path = 38 Jr, path : 102 i iii_i_!i_ii!ili!!lI.Number of II'J:53452 II :....:.., ,:.11 '.i!i!:!::!ll _J 5 10 1E 26 25 38 35 ,, 40 45 50 55 50 lib: °96 m lib ,_,,+: '65 78 75 rn, 322 88 i iiii::!i!::is_ _ _i_::i::!i!::ii iSi:i:i:i :,:.:.:.:.:. - _ ' ' -80 - 7s i -?El @iiiiill -E,.-, iiiiiii ;::::;:::::: ::;::::::::::::::: -ss -50i -4 _1 "40J i!i!iii !iiiii!iiiii!ii -ssl -301 :!Si;!:i: 8!:i:i8 @iiiiiiil Figure 7. Example of valley following route using normalized terrain data. :!:i:!:i:i:i 21 ORIGINAL COLOR PAGE PHOTOCPAPH :N_PA :88 :488 i:i:!:i:i:i:i:i :78684 m, 474 :53452 m, 322 ,: 396 m i!iiiii -25: -20 i -15 Figure 8. Example of valley following route using non-normalized terrain data. 23 PRECEDING PAGE BLANK NOT FILMED V v ORIGINAL PAGE COLOR PHOTOCRAPH iii!]iiiiiiiili!i :T_U ION :NAPA :80 A",' :4E:8 : ?8684 m, 474 9 : 53452 m, ,3-, _--_ : 396 m cit,., :5 t I0 15 20 25 :38 35 i 40 45 ,' I 50 , 55 60 65 78 75 ;-;8 -86 -75 ° ,:,:,:,:.:.:.:-:.: iiiiiii!iii!ii!i!i iiii!ii :i:!:!:i:!:i:i:i:i .:.:,:.:.:,:.:.:. :,:.:,:,:,:,:.:.: !iiii!iiiiii!iiii ::::::::::::::::: i:iS!:!:!:i:i.i !i_ili!i!ii!!ii!i !_iiiii!i!!iiii!i I i_!!ii!!!ii!ii!i! Figure 9. Route control using a control-point. 25 PREOEDiNG PAGE n,-,,.._, NOT F;LMED ,.--_,':.r, l m ORIGINAL COLOR PAGE PHOTOGRAPH Number of points Route distance, N_ber of points Route distance_ Number of points Route distance; Number of points 18 15 in path = km= 114.4.-3 in path = km = 34.15 in path = km = 33.74 in path = 28 25 38 I 97 97 29 29 35 48 i45 58 I PLANNER :TAU REGION :NAPA PLANNER:88 DI SPLA'r' :488 : 78820 : 74368 :98m 5.5 6.8 65 78 m, 478 m, 448 75 88 lb Figure 10. Route control using a cost function weighting factor. PI,'ECEDiNQ 27 " P,&qE _' _ _" _, _LA[%_ EOT F_LMED ORIGINAL COLOR PAGE PHOTOGRAPH m , • jq 8 25 38 35 48 45 58 5,5 58 65 78 75 88 18 25 38 35 48 45 58 55 60 65 78 75 88 P F4E (a) Figure 11. Route computed (b) displayed at 240 x 240. using (b) 240 x 240 (0.33 km) grid terrain data. (a) Displayed at 480 x 480, o ORIGINAL 8 25 38 35 4_ I I PAGE 5 l (a) Figure ]2. Rome compm_ (b) c]isp]ay_ at 160 x 160. (b) usin_ 160 x ]60 (0.5 kin) grid tcn'ain dam. (a) Disp]aycd at 480 x 480, 31 ,, ..... ..... I ORIGINAL COLOR PAGE PHOTOGRAPH '8 25 38 35 = -20i ....... ; :::::::::::::::::::::::::::::::: Figure 13. Route computed using 80 x 80 (1.0 km) grid terrain data. (a) Displayed at 480 x 480, (b) displayed at 80 x 80. w N|IoMI _On_s Ird Report Documentation 2. Government Accession No. Page 3. Recipient's Catalog No. SpeQe Adminil_silon 1. Report No. NASA TM- 102882 5. Report Date 4. Title and Subtitle A Workstation-Based Helicopters Evaluation of a Far-Field Route Planner for June 1991 Code 6. Performing Organization 7.Author(s) 8. Performing Organization Report No. David N. Warner, Jr. and Francis J. Moran A-91011 10. Work Unit No. 505-66-1 9. Performing Organization Name and Address l 11. Contract or Grant No. Ames Moffett Research Field, Center CA 94035-1000 13. Type of Report and Period Covered 12. Sponsoring Agency Name and Address Technical Administration 14. Sponsoring Memorandum Agency Code National Aeronautics DC and Space Washington, 20546-0001 15. Supplementary Notes Point of Contact: David Moffett (415) N. Warner, Field, 604-5443 CA Jr., Ames Research Center, MS 210-9, 94035-1000 or FTS 464-5443 16. Abstract Helicopter route planners. flight As missions part at very low, Nap of the Earth, Ames Research Center preflight or change algorithm, for route the distance mission altitudes planner, place a heavy workload on the pilot. aids in an This In time To aid in reducing selecting evaluation algorithm general, schedule, the overall this workload, (far-field) has been investigating a route the mission, in mission based various planner types of automated algorithm can be used to replan describes techniques, the mission. a flight from of an automated threats planning route to be flown. During the route planner requirements. programming a new route in case of unexpected of a candidate route This report and during on dynamic meets most of the requirements are to minimize within be flyable planning, and/or of available both preflight the requirements and must fuel and the deviation fuel and time. the constraints 17. Key Words (Suggested by Author(s)) 18. Distribution Statement NOE flight Mission Far-field Waypoint management planning path planning Unclassified-Unlimited Subject 20. Security Classif. (of this page) 21. No. of Pages Category 22. Price - 04 i9. Security Classif. (of this report) Unclassified NASA FORM 1626 OCT86 Unclassified 36 A03 For sale by the National Technical Information Service, Springfield, Virginia 22161