1、 44 6 2 0 2 3 6 A CTA A M A M E NTA IIVol 44 No 6Jun 2023DOI 10 12382/bgxb 2022 0095 210094、。V279+.2 A 1000-1093 2023 06-1564-12 2022-02-19 2019 Co llab o r at iv e Tas k Allo c at io n M e t h o d f o r M u lt i Tar g e t Air Gr o u n d He t e r o g e n e o u s Un m an n e d Sy s t e m FA N B oy an
2、g ZH A O Gaopeng B O Yumi ng W U Xi ang SchoolofA utomati on Nanj i ngUni versi tyofSci ence and Technol ogy Nanj i ng210094 Ji angsu Chi na A b s tr a c t To address the col l aborati ve task al l ocati on probl em ofai r-ground heterogeneous unmannedsy stemscomposed ofground unmanned vehi cl es UG
3、V and unmanned aeri alvehi cl es UA Vs f aci ngl arge ranges and mul ti pl e targets the task al l ocati on modelof ai r-ground heterogeneous unmannedsy stemsi sestabl i shed wi th the compl eti on ti me ofthe unmanned sy stemsasthe opti mi z ati on goaland theconstrai ntsofthe UA V l aunch and reco
4、very endurance and task sequence taken i n account A taskal l ocati on method f or mul ti-target ai r-ground heterogeneous unmanned sy stems i s proposed whi chcombi nesdensi typeak cl usteri ngand hy bri d parti cl e swarm opti mi z ati on al gori thm hy bri d-PSO to sol vethe task al l ocati on pr
5、obl em ofai r-ground heterogeneousunmanned sy stems so asto obtai n the gl obaltaskal l ocati on resul tsthatsati sf ythe constrai nts The proposed method i sveri f i ed bysi mul ati on experi ments and the resul ts show that the method can ef f ecti vel ysol ve the task al l ocati on probl em of ai
6、 r-groundheterogeneousunmanned sy stemsi n di f f erentoperati onalenvi ronments K e y w o r d s unmanned sy stem task al l ocati on densi typeak cl usteri ngal gori thm hy bri d parti cl e swarmopti mi z ati on al gori thm0、1。UA V、6、UA V。UGV、2。3。、4 7。NP-hard 8。M urray 9 UA V UA V UA V。H a 10 UA V U
7、A V UA V UA V UA Vs UA V。B ae 6 7。Chen 11 UA Vs UA V。Chen 12 UA Vs UA V。B anf i 13 UGVs-UA Vs UGVs-UA Vs。A rbanas 14 UGV UA V UGV UA V UA V UGV。Shi ma 15 E di son 16 UA Vs、3、。Feo-Fl ushi ng 17。Zhao 18 UA Vs UA V。Chen 19 UA Vs。UGV UA V UGV UA V。UGV UA V。UA V UGV、UA V。UA V UGV UGV UA V。1 1 2、。5 6 5 1
8、44 1 Fi g 1 M ai n i dea ofthi sstudy1 1.1。UGV UA V。UGV G=g UA Vs U=u 1 u 2 T=t 1 t 2、Ts T Ta T。2 UA V。UGV UA V t re UA V UA V UGV UA V t l UGV。UGV v g UA V v u h u。Tl and=Ta t re UA V t l Tl and。UA Vs UGV UGV UA V。2 Fi g 2 Schemati cdi agram ofcol l aborati ve operati onprocessofmul ti-targetai r-g
9、round heterogeneousunmanned sy stem1.2 1.2.1 UA Vs UA V UA Vs 1 mi n JU=u U t i Tu t j Tu 2k=1Xu k t i t j w u k t i t j 1 Tu UA V Tu=t re Ts Xu k t i t j 0 1 Xu k t i t j 0 1 UA V u t i t j k 1 0 k=1 k=2 w u k t i t j UA V u t i t j k。UA V 2 UA V。u U t i Tu t j Tu 2k=1Xu k t i t j w u k t i t j Wu
10、max 2 Wu max UA V u。UA V 3 4。u U t i Tu t j TsXu 1 t i t j=Ns 3 u U t i Tu Xu 1 t i t j=1 t j Ts 4 Ns。6 6 5 1 6 UA V 5 7 UA V UGV UA V。u U t i Ts t j Tl andXu 2 t i t j=NT 5 u UT t i Ts t j Tl andXu 2 t i t j=1 6 mi nu U T t i Tu t j Tu 2k=1Xu k t i t j w u k t i t j w g t 0 t l 7 NT UA V UT UA V UT
11、 U w g t 0 t l UGV UA V。1.2.2 UGV UGV UGV 8 mi n JG=t i Tg t j TaXg t i t j w g t i t j+Jwait 8 Tg UGV Tg=t re Ta Xg t i t j Xg t i t j 0 1 UGV t i t j 1 0 w g t i t j UGV t i t j Jwait UGV UA Vs Jwait=mi nu UT Wu Wg t 0 t l Wu UA V u Wg t 0 t l UGV。UGV 9 UGV UA V。u UT Jwait 0 9 UGV 10 11。u U t i T
12、t j TaXg t i t j=Nd 10 t j Ta u U t i TXg t i t j=1 11 Nd。12 UA V UGV。t i Ta Jt i U Jt i G 12 Jt i U UA V t i Jt i G UGV t i。2 1 20 n C1 C2 Cn。UGV UGV。UGV UGV PC=C1 C2 Cn UGV。2 Ci。Ci 1 UGV UGV t i 10 Ci 50 m UA V t i re UA Vs Pi U t i re、UA V t i l 2 7。UA V Pi U UA V mi nu U Ji u UGV Pi G UGV。Pi U P
13、i G Ci Pi=Pi U Pi G。PC Pi P=P1 P2 Pn。3 i t CC、i t CM i t UC、i t UM UA V i t GC、i t GM UGV。2.1 UA V。、UGV UA V UA V。7 6 5 1 44 3 Fi g 3 Processofthe task al l ocati on method f ormul ti-targetai r-ground heterogeneousunmanned sy stem。2.1.1 t p t q d i s t t p t q t p t q T。t p N t p t p、N t p=t q|d i
14、s t t p t q 13 t p p t p N t p t p=|N t p|14 t max=t p|t p T t q T t p t q 15 Smax=t p t q t n|d i s t t max t p d i s t t max t q d i s t t max t n 16 Ch Cm d i s t Ch Cm=mi n d i s t t p t k 17 t p Ch t k Cm。1。T t max C*i。Smax Smax M t p m i n De n s t p m i n De n s t p Smax 18 t p C*i。C*i=t max
15、t p|t p Smax。T=T C*i T m i n De n s。T C*=C*1 C*2。2。1 C*m i n Di s t Ci=C*i C*j|C*i C*j C*d i s t C*i C*j m i n Di s t 19 C=C1 C2。2.1.2、m i n De n s m i n Di s t。8 6 5 1 6 UA V。2.2。、UA Vs UGV 3 UA Vs UGV Ci t i max。2.2.1 2.2.1.1 UGV f i t n e s s=t i t j Tmaxp a t h G t i t j 20 Tmax T p a t h G t i
16、t j UGV t i t j。2.2.1.2 UA V UA V。f i t n e s s=u U(T t i Ts t re t j Tsp a t h U t i t j+p a t h U t u end t l)21 t re UA Vs p a t h U t i t j UA V u t i t j p a t h U t u end t l UA Vu t u end UA V u t u end Ts t l UA V t l Tl and。2.2.1.3 UGV UGV UA V f i t n e s s=t i Ta t re t j Tap a t h G t i
17、t j+c o s t wait 22 c o s t wait UGV UA V。2.2.2 UGV UA V 3、4、10 11。UGV 4 UA V 16 1 16 4 7 10 13 4。4 Fi g 4 A sampl e ofthe parti cl e code2.2.3。3、4、10 11。4 UA V 2 2 5。5 Fi g 5 A sampl e ofthe crossoveroperati on2.2.4。9 6 5 1 44 3、4、10 11。2 UA V 2 4 6。6 Fi g 6 A sampl e ofthe mutati on operati on2.2.
18、5 UA V UA V t u end Ts t i Tl and c o s t i l and=u UTp a t h U t u end t i 5 7 t l。UGV 9 12。2.2.6 1。2 UA V UA V 2 UA V 2 2 2 2。UGV。3 12 000 m 12 000 m 6000 m 6000 m 12 000 m 12 000m UGV 4 UA V。6000 m 6000 m 12 000 m 12 000 m。、。1000 m 5。UGV 20 km/h UA V 30 km/h 30 mi n。A MD y z enTM7 P O 4750G 3.6GH
19、 z 32 GB Ubuntu 20.04。3.1 12 000 m 12 000 m 50 7。7 Fi g 7 esul tofdensi typeak cl usteri ng Davi es-B oul di n DB I 0 7 5 1 6 DB I=1N Ni=1maxj i d i+d j d i j 23 N d i Ci d j Cj d i j Ci Cj。DB I。1 12 000 m 12 000 m、DB SCA N k-means。、DB SCA N k-means。、。1 12 000 m 12 000 m Tabl e 1 Compari son ofcl us
20、teri ngresul tsi n the range of12 000 m 12 000 m DB I 16 0.9281DB SCA N 16 1.0168k-means 16 1.58593.2 6000 m 6000 m 8 50 17 8 6000 m 6000 m Fi g 8 Targetdi stri buti on map i n the range of6000 m 6000 m。UGV。2。3 9。UA Vs 19 50m 365 m 106m 40 UGV UA Vs Jwait=5.36s UA V 1718.11s 1340.85 s 1562.53s 1643.
21、87s UA Vs 2、3 4、5 7、UGV 9、10 11 12。2 6000 m 6000 m Tabl e 2 Parametersofthecol l aborati ve task al l ocati on method i n the range of6000 m 6000 m/m 1000 5/m 1500 UA V UGV 800 250 2 400 800 0.7 0.7 0.7 0.7 3 6000 m 6000 m Tabl e 3 Col l aborati ve task al l ocati on i n the range of6000 m 6000 m/m/
22、s UGV/sUA V1 7 38 10 4 44 42 23 36 12 33 49 29 1718.11UA V2 11 15 8 5 14 20 28 48 46 22 1340.85UA V3 365 106 47 9 21 34 1 39 31 16 32 3 50 27 40 6 1562.53 40 1335.49UA V4 19 45 18 17 30 26 24 43 37 13 41 35 2 25 1643.87UGV 7 48 22 40 5 45 11 8 2 35 41 13 30 16 44 23 49 4323.731 7 5 1 44 9 6000 m 600
23、0 m Fi g 9 Schemati cdi agram ofthe col l aborati ve taskal l ocati oni n the range of6000 m 6000 m 9。UA V 10 555 2 1800。859.9s。10 UA V Fi g 10 Fl i ghtti me curvesofUA Vs3.3 12 000 m 12 000 m 11 50 17。UGV。4。5 UA V 3401.53 s 2 884.75 s 2 491.79s 2 704.78 s UA Vs 2。11 12 000 m 12 000 m Fi g 11 Target
24、di stri buti on map i n the range of12 000 m 12 000 m 4 12 000 m 12 000 m Tabl e 4 Parametersofthe task al l ocati on method wi thoutcl usteri ngi n the range of12 000 m 12 000 m UA V UGV 800 250 2 400 800 0.7 0.7 0.7 0.7 5 12 000 m 12 000 m Tabl e 5 Task al l ocati on resul twi thoutcl usteri ngi n
25、 therange of12 000 m 12 000 m/m UA V1 14 37 38 30 50 49 20 26 6 36 17 33 45 5 UA V2 617 85 1 24 18 3 47 34 25 19 15 23 UA V3 41 35 46 8 40 9 16 39 4 44 31 7 13 27 UA V4 2 10 28 48 22 12 43 29 32 42 11 5 21 6。16 12。2 7 5 1 6 6 12 000 m 12 000 m Tabl e 6 Parametersofthe col l aborati ve task al l ocat
26、i on method i n the range of12 000 m 12 000 m/m 1000 5/m 1500 UA V UGV 240 240 50 720 720 200 0.7 0.7 0.7 0.7 0.7 0.7 12 Fi g 12 Cl usteri ngand task al l ocati on 7 13。3 5 35 29 UGV UA V J3wait=92.9731s、J16wait=1.41s UGV UA Vs。UA V 1 584.87 s 1661.48 s 1737.412 5 s 1488.49s UA Vs 2、3 4、5 7、UGV 9、10
27、 11 12。7 12 000 m 12 000 m Tabl e 7 Col l aborati ve task al l ocati on f ormul ti-targetai r-ground heterogeneousunmannedsy stem i n the range of12 000 m 12 000 m/m/s UGV/s1UA V3UGV 617 85 2 2 23.933935.9009 02 UA V1 3846 642 10 41 28 185.7957 03UA V2UGV 5700 1631 35 46 35 110.946117.97 30 35 17.97
28、304UA V1UA V4UGV 6843 3829 8 48 22 40 16 9 8 458.0001696.072 035.9460 05UA V1 9296 4086 24 37 519.317 5UA V2 12 29 6 26 20 1020.022 1UA V3 18 43 36 32 531.2584 29 517.9066UA V4 1 14 792.6202UGV 18 43 29 26 32 12 14 2 815.35356 UA V3 503 6752 38 23.9832 0/m/s UGV/s7UA V2UGV 1233 11174 30 50 30 181.64
29、1735.8702 08 UA V1 3787 11664 49 23.8713 09UA V2UGV 6727 10400 42 42 23.954435.9315 010 UA V1 7 091 8449 11 24.1054 011UA V2UGV 8983 8013 5 21 47 3 5 21 373.783992.712 1 012UA V2UGV 7 940 6654 17 33 45 45 276.8789390.8494 013 UA V2 9218 4227 39 24.0539 014UA V3UGV 10211 5896 7 31 4 44 27 13 27 660.2
30、844532.2189 015 UA V2 11789 9162 34 23.982 0 016UA V3UGV 9868 11695 25 19 15 23 23 497.952 6510.7207 03 7 5 1 44 13 12 000 m 12 000 m Fi g 13 Schemati cdi agram ofthe col l aborati ve task al l oca-ti on f or mul ti-target ai r-ground heterogeneous un-manned sy stem i n the range of12 000 m 12 000 m
31、 13。277.93s。4、。1。2 UA Vs、。e f e r e n c e s 1 J 2020 46 13 1 20 W U W L ZH OU XS SH E N B etal A revi ew ofswarm roboti csy stemspropertyeval uati on J A cta A utomati ca Si ni ca 2020 46 13 1 20 i n Chi nese 2 DE NG Q B YU J Q W A NG N F Cooperati ve task assi gnmentofmul ti pl e heterogeneousunman
32、ned aeri alvehi cl esusi ng a modi f i edgeneti cal gori thm wi th mul ti-ty pe genes J Chi nese JournalofA eronauti cs 2013 26 5 1238 1250 3 GA ONE E NA LDI CA SA VOLA A et al Cooperati vemi ssi on pl anni ng f or a cl ass of carri er-vehi cl e sy stems C Proceedi ngsofthe 49thIE E E Conf erence on
33、Deci si onand Control Pi scataway NJ US IE E E 2010 1354 1359 4 W U W N W A NG XG CUI N G Fast and coupl ed sol uti on f orcooperati ve mi ssi on pl anni ng ofmul ti pl e heterogeneous unmannedaeri alvehi cl es J A erospace Sci ence and Technol ogy 2018 79 131 144 5 CH E N M Q CH E N Y CH E N ZH eta
34、l Path pl anni ngofUA V-UGV heterogeneous robot sy stem i n road network C Proceedi ngsofInternati onalConf erence on Intel l i gent oboti csandA ppl i cati ons Cham Swi tz erl and Spri nger 2019 497 507 6 B A E J PA K M A heuri sti cf oref f i ci entcoordi nati on ofmul ti pl eheterogeneous mobi l
35、e robots consi deri ng workl oad bal ance J IE E E oboti cs and A utomati on Letters 2021 6 2 4064 4070 7 B A E J LE E J CH UNG W A heuri sti cf or task al l ocati on androuti ng ofheterogeneous robots whi l e mi ni mi z i ng maxi mum travelcost C Proceedi ngs of 2019 Internati onal Conf erence on o
36、boti cs and A utomati on M ontreal Canada IE E E 2019 4531 4537 8 J 2021 42 7 1482 1495 ZH A NG Y F LIN D F ZH E NG D et al Task al l ocati on andtraj ectory opti mi z ati on of UA V f or mul ti-target ti me-spacesy nchroni z ati on cooperati ve attack J A cta A rmamentari i 2021 42 7 1482 1495 i n
37、Chi nese 9 M U A Y C C CH U A G The f l y i ngsi deki ck travel i ngsal esmanprobl em Opti mi z ati on of drone-assi sted parcel del i very J Transportati on esearch Part C E mergi ng Technol ogi es 2015 54 86 109 10 H A Q M DE VILLE Y PH A M Q D et al On the mi n-costtravel i ng sal esman probl em
38、wi th drone J Transportati on esearch PartC E mergi ngTechnol ogi es 2018 86 597 621 11 CH E N J C DU C L LU X et al M ul ti-regi on coverage pathpl anni ngf orheterogeneousunmanned aeri alvehi cl essy stems C Proceedi ngsof2019 IE E E Internati onalConf erence on Servi ce-Ori ented Sy stem E ngi ne
39、eri ng San Franci sco B ay CA US IE E E 2019 356 365 12 CH E N J C DU C L ZH A NG Y et al A cl usteri ng-basedcoverage path pl anni ng method f or autonomous heterogeneousUA Vs J IE E E Transacti ons on Intel l i gent Transportati onSy stems 2021 23 12 25546 25556 13 B A NFI J M E SSING A K ONINGE C
40、 et al H i erarchi calpl anni ng f or heterogeneous mul ti-robot routi ng probl ems vi al earned subteamperf ormance J IE E E oboti csand A utomati onLetters 2022 7 2 4464 4471 14 A B A NA S B IVA NOVIC A CA M et al Decentral i z ed4 7 5 1 6 pl anni ng and controlf or UA V-UGV cooperati ve teams J A
41、 utonomous obots 2018 42 8 1601 1618 15 SH IM A T A SM USSE N S J SPA KS A G et al M ul ti pl etask assi gnmentsf orcooperati nguni nhabi ted aeri alvehi cl esusi nggeneti c al gori thms J Computers Operati ons esearch 2006 33 11 3252 3269 16 E DISON E SH IM A T Integrated task assi gnment and patho
42、pti mi z ati on f or cooperati ng uni nhabi ted aeri alvehi cl es usi nggeneti c al gori thms J Computers Operati ons esearch 2011 38 1 340 356 17 FE O-FLUSH ING E GA M B A DE LLA L M DI CA O G A Spati al l y-di stri buted mi ssi ons wi th heterogeneous mul ti-robotteams J IE E E A ccess 2021 9 6732
43、7 67348 18 ZH A O XY ZONG Q TIA N B L etal Fasttask al l ocati on f orheterogeneous unmanned aeri alvehi cl es through rei nf orcementl earni ng J A erospace Sci ence and Technol ogy 2019 92 588 594 19 CH E N L LIU W L ZH ONG J A n ef f i ci ent mul ti-obj ecti ve antcol onyopti mi z ati on f ortask
44、 al l ocati on ofheterogeneousunmannedaeri alvehi cl es J JournalofComputati onalSci ence 2022 58 101545 20 J 2009 26 4 539 548 W A NG J XIA L N JING J W M axi mum densi tycl usteri ngal gori thm J Journalof Uni versi tyof Chi nese A cademyofSci ences 2009 26 4 539 548 i n Chi nese 1991。E-mai l f anboy ang nj ust edu cn 1983、。E-mai l z haogaopeng nj ust edu cn 1965。E-mai l by mi ng nj ust edu cn 1989。E-mai l wux0213 hotmai l com5 7 5 1
