Fechar
Metadados

@InProceedings{AmaralLopeHéli:2007:NoLiKa,
               author = "Amaral, Jairo Cavalcanti and Lopes, Oliveira e Souza Marcelo and 
                         H{\'e}lio, Koiti Kuga",
                title = "A Non linear Kalman Filter Algorithm for Monitoring the Motion of 
                         a Vehicle in Real Time Around a Reference Trajectory",
            booktitle = "Proceedings...",
                 year = "2007",
         organization = "SAE BRASIL , 16. International Congress and Exposition.",
            publisher = "SAE International",
             abstract = "In this work we present a nonlinear Kalman filter algorithm for 
                         monitoring the motion of a vehicle in real time around a reference 
                         trajectory. It is based on the linearized and/or the extended 
                         Kalman filters. The first vehicle and reference trajectory 
                         considered are a satellite and a keplerian orbit around Earth. The 
                         orbital motion is modeled as simple as possible so as not to 
                         burden the computer load and, at the same time, to provide enough 
                         accuracy to this sort of problem. The source of measurements is 
                         simulated to represent actual measurements such as GPS, 
                         geo-stationary satellite to satellite tracking, or conventional 
                         ground tracking stations. Some tests are done to show its 
                         correctness. After that, we intend to extend such algorithm to 
                         other vehicles and reference trajectories as: 1) a triangular 
                         equatorial constellation of satellites orbiting around Earth in 
                         coplanar keplerian orbits; 2) a semi-autonomous car moving in a 
                         sensored track; 3) etc. The accuracy in motion estimation will 
                         ultimately affect the performance of controlling and maneuvering 
                         the vehicle, therefore providing a basis for its design, 
                         implementation and maintenance. RESUMO: Neste trabalho 
                         apresentamos um algoritmo com filtro de Kalman n{\~a}o linear 
                         para o monitoramento de um ve{\'{\i}}culo em tempo real em torno 
                         de uma trajet{\'o}ria de refer{\^e}ncia. O primeiro 
                         ve{\'{\i}}culo e trajet{\'o}ria considerados s{\~a}o um 
                         sat{\'e}lite e uma {\'o}rbita kepleriana em torno da Terra. O 
                         movimento orbital ser{\'a} modelado o mais simples 
                         poss{\'{\i}}vel para n{\~a}o prejudicar a carga computacional 
                         e, ao mesmo tempo, fornecer precis{\~a}o suficiente para este 
                         tipo de problema. A fonte de medidas ser{\'a} simulada para 
                         representar medidas reais, tais como GPS, rastreamento via 
                         sat{\'e}lite, ou esta{\c{c}}{\~o}es convencionais de 
                         rastreamento. Alguns testes s{\~a}o feitos para mostrar sua 
                         corre{\c{c}}{\~a}o. Ap{\'o}s isso, pretendemos estender tal 
                         algoritmo para outros ve{\'{\i}}culos e trajet{\'o}rias de 
                         refer{\^e}ncia tais como: 1) uma constela{\c{c}}{\~a}o 
                         equatorial triangular de sat{\'e}lites orbitando a Terra em 
                         {\'o}rbitas keplerianas coplanares; 2) um carro 
                         semi-aut{\^o}nomo movendo-se numa pista sensoriada; 3) etc. A 
                         precis{\~a}o da estima{\c{c}}{\~a}o de movimento ir{\'a} 
                         afetar a performance de controle e de manobra do ve{\'{\i}}culo, 
                         fornecendo assim uma base para seu projeto, 
                         implementa{\c{c}}{\~a}o e manuten{\c{c}}{\~a}o.",
  conference-location = "Warrendale",
                  doi = "10.4271/2007-01-2682",
                  url = "http://dx.doi.org/10.4271/2007-01-2682",
             language = "en",
        urlaccessdate = "21 jan. 2021"
}


Fechar