[{"data":1,"prerenderedAt":160},["ShallowReactive",2],{"method-fastslam2002":3},{"method":4,"reference":46,"equipment":66,"figures":76,"results":77},{"id":5,"label":6,"shortName":7,"title":8,"year":9,"era":10,"cluster":11,"scope":12,"keyIdeaZh":13,"keyIdeaEn":14,"fulltextStatus":15,"publicationStatus":16,"recommendation":17,"constructionRelevance":18,"validationEnvironment":19,"strengths":23,"limitations":27,"sensors":30,"platform":32,"estimator":34,"association":35,"timeModel":36,"deskew":37,"loopClosure":38,"globalOptimization":39,"mapRepresentation":40,"prior":39,"outputGeometry":41,"compute":42,"codeUrl":43,"codeLicense":44,"relatedVersions":45},"fastslam2002","Montemerlo et al., 2002","FastSLAM","FastSLAM: A Factored Solution to the Simultaneous Localization and Mapping Problem",2002,"classic","C01","full_slam_with_global_correction","FastSLAM 利用「給定機器人路徑時各地標條件獨立」的性質，把 SLAM 後驗分解為路徑分布與各地標的條件分布：以粒子濾波器（particle filter）取樣路徑，每個粒子再為每個地標維持一個小型 EKF。作者以樹狀資料結構使每次更新的時間複雜度降為 O(M log K)，並讓每個粒子各自做資料關聯（data association），因此可同時追蹤多種關聯假設。模擬中地標數擴充到 50,000 個，實體機器人實驗則以人工測得的地標位置比對。","Factors the SLAM posterior into a particle-sampled path and independent per-landmark EKFs, giving O(M log K) updates and per-particle data association.","full_text_reviewed","peer_reviewed_published","background","未在營建場域測試（NASA 火星車研究用小型場地與模擬）。屬 2D 地標式 SLAM 背景知識，不能直接代替 3D 工程點雲比較。",[20,21,22],"simulation","controlled_experiment","independent_reference",[24,25,26],"Scales logarithmically with the number of landmarks","simulated maps up to 50,000 landmarks with 100 particles, using about 0.3% of the parameters of a conventional EKF (abstract","Experimental Results) | A fixed particle count (e.g., M = 100) appeared to work well across a large number of situations, and more landmarks mildly reduced pose and map error (Experimental Results, Fig. 6) | Per-particle data association can recover from wrong associations more readily than a single EKF hypothesis (Data Association section)",[28,29],"The authors report that in some situations the number of particles required for accurate mapping may be prohibitively large (Introduction).","Stachniss et al. (Handbook 2016, Sec. 46.2.4) note that the number of particles can grow very large for nested loops.",[31],"[\"2D laser range finder (SICK)\",\"robot controls u_t (odometry sensor not named in the paper)\"]",[33,20],"wheeled UGV","Rao-Blackwellized particle filter (particles over robot path, one small EKF per landmark per particle)","per-particle maximum-likelihood landmark association with new-landmark threshold","discrete poses","not_reported","implicit through particle weighting and resampling; no explicit loop-closure module","none","point landmarks stored in a balanced binary tree per particle","landmark positions and robot path per particle","O(M log K) per update with M particles and K landmarks (Efficient Implementation section)",null,"not_applicable",[],{"id":5,"kind":47,"shortName":7,"title":8,"authors":48,"year":9,"venue":53,"venueType":54,"publisher":55,"volumeIssuePages":56,"doi":43,"arxivId":43,"url":57,"firstPublicDate":58,"publicationStatus":16,"metadataStatus":59,"fulltextStatus":15,"era":10,"classicReason":60,"codeUrl":43,"cluster":11,"topics":61,"mdpi":62,"verification":63,"label":6,"fulltextRoute":64,"versionRead":65,"addedByCensus":62},"method",[49,50,51,52],"Michael Montemerlo","Sebastian Thrun","Daphne Koller","Ben Wegbreit","Proceedings of the Eighteenth National Conference on Artificial Intelligence (AAAI-02)","conference","AAAI","pp. 593-598","https:\u002F\u002Fcdn.aaai.org\u002FAAAI\u002F2002\u002FAAAI02-089.pdf","2002","metadata_verified","necessary technical node: introduces the Rao-Blackwellized factorization (particle filter over paths, per-particle landmark EKFs) that later grid-based RBPF mappers such as GMapping build on.",[11],false,"confirmed","publisher OA","AAAI-02 proceedings paper, pp. 593-598, publisher-hosted PDF (cdn.aaai.org), 6 pages; no DOI exists",[67,72],{"category":68,"model":69,"canonical":69,"role":70,"dataset":43,"specs":37,"locator":71},"platform","Pioneer robot","method input","Experimental Results, Fig. 4",{"category":73,"model":74,"canonical":74,"role":70,"dataset":43,"specs":37,"locator":75},"lidar","SICK laser range finder","Experimental Results",[],{"totalRows":78,"groupCount":79,"groups":80,"others":159},3,2,[81,122],{"slug":82,"group":83,"sourceId":5,"sourceLabel":6,"table":84,"selfRows":79,"metrics":85,"seqs":93,"entrants":100,"cells":106,"outcomes":114,"locators":115,"hardware":117,"wordings":118,"notes":119},"fastslam2002-text-experimental-results","fastslam2002:Text Experimental Results","Text Experimental Results",[86,90],{"label":87,"unit":88,"statistic":89,"alignment":37},"average residual map error compared to the manually generated map","cm","mean",{"label":91,"unit":92,"statistic":37,"alignment":39},"number of parameters relative to the conventional EKF","%",[94,98],{"dataset":95,"sequence":96,"environment":97},"NASA-funded Mars rover test arena","single straight-line run","small test arena set up for Mars rover research (indoor or outdoor not stated)",{"dataset":20,"sequence":99,"environment":20},"50,000 landmarks",[101,104],{"name":102,"methodId":5,"linkable":103,"proposed":103,"self":103},"FastSLAM (M = 10 samples)",true,{"name":105,"methodId":5,"linkable":103,"proposed":103,"self":103},"FastSLAM (M = 100)",[107,111],[108,108,108,109,110,108,110,110,108],0,8.3,-1,[112,112,112,113,110,112,110,110,112],1,0.3,[],[116,75],"Experimental Results, Fig. 4c",[],[],[120,121],"Physical testbed: Pioneer robot with SICK laser mapping rocks in a Mars-rover research arena; FastSLAM map compared with manually determined landmark (rock) locations","Simulation with 50,000 landmarks mapped with as few as 100 particles",{"slug":123,"group":124,"sourceId":125,"sourceLabel":126,"table":127,"selfRows":112,"metrics":128,"seqs":132,"entrants":137,"cells":144,"outcomes":151,"locators":152,"hardware":154,"wordings":156,"notes":157},"fastslam2-2003-table-sec-6","fastslam2_2003:Table Sec. 6","fastslam2_2003","Montemerlo et al., 2003","Table Sec. 6",[129],{"label":130,"unit":131,"statistic":37,"alignment":39},"processing time for the whole data set","s",[133],{"dataset":134,"sequence":135,"environment":136},"Victoria Park (Sydney)","full data set (3.5 km)","outdoor park",[138,140,142],{"name":139,"methodId":43,"linkable":62,"proposed":62,"self":62},"EKF",{"name":141,"methodId":5,"linkable":103,"proposed":62,"self":103},"regular FastSLAM, M=50 particles",{"name":143,"methodId":125,"linkable":103,"proposed":103,"self":62},"FastSLAM 2.0, M=1 particle",[145,147,149],[108,108,108,146,110,108,108,110,108],7807,[112,108,108,148,110,108,108,110,108],315,[79,108,108,150,110,108,108,110,108],54,[],[153],"Sec. 6 table",[155],"1GHz Pentium PC",[],[158],"Total time to process the Victoria Park data set on a 1 GHz Pentium PC; data acquisition took 1,550 s",[],1790510661304]