This thesis focuses on the problem of SLAM, in particular feature-based SLAM, where a mobile autonomous vehicle explores an unknown environment, navigating by observing landmarks. As the environment is previously unexplored, SLAM algorithms must estimate the location of the landmarks in addition to the location of the vehicle. The majority of research in feature-based SLAM builds on the legacy of foundational work using the Extended Kalman Filter (EKF), a single-object estimation technique. Because feature-based SLAM is an inherently multi-object problem, this has led to a number of suboptimalities in popular solutions. We hypothesize that a feature based SLAM algorithm derived from multi-object estimation techniques can achieve superior estimation performance compared to conventional algorithms, especially in conditions where the aforementioned weaknesses are most promi-
\nnent. In this work, we develop such an algorithm, using the SC-PHD filter, a multi-object estimator modeled on cluster processes. This algorithm hosts capabilities not typically seen with feature-base SLAM solutions such as principled handling of false alarm measurements and missed detections, and navigation with a mixture of stationary and moving landmarks. We
\npresent experiments with the SC-PHD SLAM algorithm on both synthetic and real datasets using an autonomous underwater vehicle. We also compare our method to the RB-PHD SLAM, showing that it requires fewer approximations in its derivation and thus achieves superior performance.<\/p>\n
<\/p>\n
__________________<\/p>\n
<\/p>\n
Aquesta tesis es centra en el problema de la localitzacio\u0301 y la construccio\u0301 de mapes de forma simulta\u0300nia (SLAM), en particular en el SLAM basat en caracter\u0131\u0301stiques, on un vehicle mo\u0300bil auto\u0300nom explora un entorn desconegut navegant mentre observa aquestes caracter\u0131\u0301stiques. Donat que inicialment l\u2019entorn es desconegut, els algoritmes SLAM han d\u2019estimar la localitzacio\u0301 d\u2019aquestes caracter\u0131\u0301stiques de l\u2019entorn al mateix temps que estima la localitzacio\u0301 del propi vehicle. La major part de la recerca en SLAM basat en caracter\u0131\u0301stiques s\u2019alimenta del llegat fonamental en filtres de Kalman este\u0300s (EKF), te\u0300cnica d\u2019estimacio\u0301 uni-objecte. Al ser el SLAM basat en caracter\u0131\u0301stiques intr\u0131\u0301nsecament un problema multi-objecte s\u2019han popularitzat un nombre de solucions sub-o\u0300ptimas. En aquesta tesis en fa la hipo\u0300tesi de que un SLAM basat en caracter\u0131\u0301stiques derivat d\u2019una te\u0300cnica d\u2019estimacio\u0301 multi-objecte pot aconseguir un major grau de rendiment en l\u2019estimacio\u0301 comparat amb els algoritmes convencionals, especialment en aquelles situacions on les debilitats anomenades anteriorment so\u0301n me\u0301s presents. En aquesta tesis es desenvolupa aquest algoritme a partir d\u2019un filtre PHD amb un u\u0301nic grup (SC-PHD), una te\u0300cnica d\u2019estimacio\u0301 multi-objecte basat en processos d\u2019agrupacio\u0301. Aquest algoritme te\u0301 unes capacitats que normalment no es veuen en els algoritmes de SLAM basats en caracter\u0131\u0301stiques, ja que e\u0301s capac\u0327 de tractar falses caracter\u0131\u0301stiques, aix\u0131\u0301 com caracter\u0131\u0301stiques no detectades pels sensors del vehicle, a me\u0301s de navegar en un entorn amb la prese\u0300ncia de caracter\u0131\u0301stiques esta\u0300tiques i caracter\u0131\u0301stiques en moviment de forma simulta\u0300nia. Es presenten els resultats experimentals de l\u2019algoritme SC-PHD en entorns reals i simulats utilitzant un vehicle auto\u0300nom submar\u0131\u0301. Els resultats so\u0301n comparats amb l\u2019algoritme de SLAM Rao- Blackwellized PHD (RB-PHD), demostrant que es requereixen menys aproximacions en la seva derivacio\u0301 i en consequ\u0308e\u0300ncia s\u2019obte\u0301 un rendiment superior.<\/p>\n","protected":false},"excerpt":{"rendered":"