未知环境中移动机器人并发建图与定位_CML_的研究进展

»26 »4ù2004M7机器人 ROBOTVol.26,No.4July, 2004ÓcI|: 1002-0446(2004)04-0380-05éÌâÏMî¦i?ymÐçÊ(CML)¥ùîÉZX王 璐, 蔡自兴(Ï2vА SÐÐýñÐýÆ?"dÐÈqùîî,µ2Éì 410083)K 1:8ÍM @›¥CMLZE,§×1 ò19Ð9Ty/¹m¥Vñ[#Â…) Ø‚’çaÂ…VU¹m.ÎCMLÙ5¥4ÄÉ›s,i)à Ÿ¥ùît].1oM: CML/SLAM;ZE;4Ä;t]Ïms Ë|: TP24 ÓDSM’: AProgress of CML for Mobile Robots in Unknown EnvironmentsWANGLu, CAIZ-ixing( Institute of Intelligent System and Software, School of Information Scienceand Engineering, Central South University, Changsha 410083, China)Abstract: ThispapersketchessomepopularmethodsofCML(ConcurrentMappingandLocalization) inrecentyears, putsem-phasisoncomparingtheirprocessesofestimationandincrementalmapping, andwaystodealwithuncertaininformationandtorep-resentmaps. ThedifficultpointsofCMLareanalyzed, andthetrendsofresearchinthefuturearediscussed.Keywords: CML/SLAM; methods; difficulties; trends1 „ý(Introduction)V LçÊ^1ö¦LC½†t¥5 %Hq[1].5¹m¥·/,¦ V[‚nž1&ÊÂ,LCÚ’çÊ.ŒÃ©ÌâÏ,¦›†Àµºµ“a“‚›œ¥Ìâ©M,¦Ìâ¥ªMº?YVîz{¥.Ë,Â^ha; Âra2,©,¤M1,ÜV) Ø|µry/Ìâ¹m,i NÉ›çÊ.ƒÕf ƒ/,ymÐçÊ^Bñ/5µÎ^5µ‘0Ù5:¹LCÚ’çÊ,¦³1 æ¨ÊÂM’ç¥Ìâ+˜(¹m)‚nž1&ÊÂ;7¹’çÌâ+˜¥ÊÂ,¦³1³1&î¥Ú’ÊÂ.CML(ConcurrentMapping and Localization)/SLAM(SimultaneousLocalization andMapBuilding)Ù5 V[퍹:¦Ã©ÌâÏVBñéʠ7SMî,MîVñÏô ÊÂ9„.Ë” É›1&çÊ,]Hy/9T¹m.çÊÐ9Tymµ¹B8,7‚^ÿ ë¥ñ¨.CMLZEV€¥E+MÏïvî¹¦tÙ5¥ùî£Ä,„v¥ùî¦ô,i|¤“L¨Ÿ¥îT,^Ž !i?ymÐçÊ¥? ï$¸¦ª¹^¦^Ž?LC1ö¥1o¥-4Hq[2].'ÓÍ+M Ÿ@›¥CMLZEÉ›9²as Ë„1 ,·Bt$€³ %¥1oÙ5,iYVI³X|¤¥îT·à Ÿ¥ùîZ_.2 CMLZEeº(Brief description of CML)SmithRaSelfMaCheesemanP¿1986M4¿EKF(ExtendedKalman Filter)¥StochasticMappingZE[3],£ 7CMLù.[ª¥E+MÏ,ùîS¶‚ ”v:Vµ¦ý^Sž›†1öaV¼=ž¼“,C¸CMLZE.®¿1ö¦%µ¥+Ä(ŒC1&Ê„Ìâ¥5; L“†„=†.Ë¤©M;Ìâa.Ë#¦î'& µ‚’çŸ),òÕCMLZEB, Ÿû^Bñ/9)nž0¥Vñ,A¶³ %/Ù5:XÁ[":SE1– SÐÁ'ù["(60234030).là°ù:2003- 04- 20 #Â…9¦#Ìâ+˜ÿ?#Â…nž„÷¹m?#¹mÂ…VU?#‚’çŸÂ…) Ø?[/Í+M@›¥CMLZEÉ›e1¥9²„1 .2. 1 ¿EKF¥CML/ SLAMEKFa†³ %dLŸ"d¥9Ù5.ƒÕCML¨ÜëUSVU¦„Ìâ+˜¥ÊÂ,|¦îÐÌâ+˜¥1"퍹ñdLŸ˜:¦î˜„4©˜.Xk = f(Xk-1, Uk) + wk (1)Zk = h(Xk) + vk (2)wk„vksYVU¦î‚’矄.Ë4©‚’çŸ,iL!¹ (´¹0¥žÿâ.2.®¿SmithR©ª¹‚’çŸ^ bW1"=¥%µŸÉ,yNÌâ¹m‚Ç “¦„Ìâ+˜¥US,Î “Bñí¦ÐÌâ+˜­W1"[#Õ‚’知xZµ”.¹mFî “:X = (XR, Xf1, , , Xfn)T (3)C(Xk) = CRR CRfCfR Cff(4)C(Xk)¥„Æ^EKFZEKv¥+Ä,ñ¥9Ø^EKFØE¥‚. C(Xk)Û"9nž¥‚É›†Ø¾h,y7žÈf ƒ¾ØEA–l û.N“, C(Xk)L=c†s:"dÿ­W¥xZµ„4©.2¥xZµ.Ï"dÿ­W¥xZµ v,5VU"dʵ v,‹P¨4©´Ð V?¹nž;Ï4©.2¥xZµ v,5VU4©” ‚V L,ºP¨4©´S^±nž.‹¨î˜„4©˜,¿EKF¥ymÐçÊ V[B,¹BñÖÌÝ}¥9)nžVñ:n5YV˜(1)9¦¥ÊÂ,iYV(2)9 V?4©¥Ìâ+˜,–ª9ØL=4©„94©W¥µ,8†"dxZµ9Ø 5:£ro•”K,i¨K-ë9¥¦ÊÂÉ›nž,Kª|4©¥Ìâ+˜Fƹm.¦MîVñÏÖÌ‚¹9)nž,Ðh" Ƶ,¤žÐ V?’¥çʐ,£t¥¨ æ›î.EKFØE) Ø‚’çZëµÿž­),yNsž¦t5×¥r,÷‚AªîÿÌâÏ¥Ì⥂’çŸ.Â…) Ø‚’çŸ,ÐhòÕ Z4i¥‚’矄¥µ^CML¥1o,9^òÕØE¥ÚÝ.3.2 ” 1 ó” 1 ó· 樒-.Ë©ž¥mñ4©´¹mÏXÜi¥nñ+˜É›÷H,A¶ü’·ñ4©´‹¿'ñ+˜,^Bñ+˜^Bñ.2” .­¿EKF Ÿª,” 1ó‚’|ÁØE?Û..d¥” 1 óZE¨NN(nearestneighbor),¾ZEe†,Œ  Ö“Í¥ñ+˜,ØE“¸^>r.ÍM1¿” 1 ó¥ùîïv9F,M?4¿ ó†Y¸Ÿ©k¥” 1 óZE[20]a¿+…1"¥” 1 óZE[21].7¿Àq¥ZEFó¹¡V” 1 óÙ5,ŒdÈÚÅ¥9Ø}N‚ æ¿v?Ì⥋¨.3.3 ØE¯‡ž›†1ö¥¦1pymÐçÊA¶?L9Ø,y79د¥1p^dÈ N Y¥.C¥ØEK?›+?+ñ’ù, Vi%„9Ø¥Ìâ+˜rž104),Œ¯–ЇLW¸µ“vµ .†VØΘ,Cµ9ØÐ¦M1Ù–µµ ,ŒƒÕµ ï°vhlt].ŒÉ›çÊ‘©) ØH,9Øùù‚?Ð¦M1.ðy­Bü^¦7¥i›9Ø#sƒTi%.|*Ü©Ž/Œñi›ØE‹¨¿CML,µ V? v–4Ú。