Abstract:Resampling process often causes the"sample impoverishment"problem in FastSLAM 2.0.In order to improve the algorithm performance and to increase the estimation accuracy,FastSLAM 2.0 is combined with genetic algorithm,and a solution named"Genetic FastSLAM 2.0"is presented for the SLAM problem.Based on the specialty of FastSLAM 2.0, an improved genetic algorithm is designed with attention to both the particle weight and the samples’ diversity.Genetic FastSLAM 2.0 estimates the robot path with unscented particle filter (UPF),and the map with extended Kalman filter (EKF). Experiments are carded out with a benchmark dataset named"car park dataset"to evaluate performance of the genetic Fast- SLAM 2.0,and the results indicate that the genetic FastSLAM 2.0 performs well on both estimation accuracy and consistency, and the computational complexity satisfies the requirements from real-time applications.
[1] Smith R C,Cheesman P.On the representation and estimation of spatial uncertainty[J].The International Journal of Robotics Research,1986,5(4):56~68.
[2] Durrant-Whyte H E Uncertain geometry in robotics[J].IEEE Journal of Robotics and Automation,1988,4(1):23~31.
[3] Smith R C,Self M,Cheeseman P.Estimating uncertain spatial relationships in robotics[A].Autonomous Robot Vehicles[M].New York,USA:Springer-Verlag,1990.167~193.
[4] Thrun S,Burgard W,Fox D.A probabilistic approach to concurrent mapping and localization for mobile robots[J].Machine Learning,1998,31(1-3):29~53.
[5] Guivant J E,Nebot E M.Optimization of the simultaneous localization and map-building algorithm for real-time implementation[J].IEEE Transactions on Robotics and Automation,2001,17(3):242~257.
[6] Doucet A,de Freitas J,Murphy K,et al.Rao-BIackwellized particle filtering for dynamic Bayesian networks[A].Proceedings of the Conference on Uncertainty in Artificial Intelligence[C].San Fransisco,CA,USA:Morgan Kaufmann,2000.176~183.
[7] Montemerlo M,Thrun S,Koller S T D,et al.FastSLAM 2.0:An improved particle filtering algorithm for simultaneous localization and mapping that provably converges[A].Proceedings of the International Conference on Artificial Intelligence[C].California,CA,USA:LICAI,2003.1151~1156.
[8] Kim C,Sakthivel R,Chung W K.Unscented FastSLAM:A robust algorithm for the simultaneous localization and mapping problem[A].Proceedings of the IEEE International Conference on Robotics and Automation[C].Piscataway,NJ,USA:IEEE,2007.2439~2445.
[9] van der Merwe R,de Freitas N,Doucet A,et al.The Unscented Particle Filter[R].Portland,OR,USA:Oregon Grnd.uate Institute,2000.
[10] Julier S J,Uhlmann J K.A new extension of the Kalman filter to nonlinear systems[A].Proceedings of the SPIE (vol.3068)[C].Bellingham,WA,USA:SP1E,1997.182~193.
[11] Doucet A,Godsill S,Andrieu C.On sequential Monte Carlo sampling methods for Bayesian filtering[J].Statistics and Computing,2000,10(3):197~208.
[12] Arulampalam M S,Maskell S,Gordon N,et al.A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking[J].IEEE Transactions on Signal Processing,2002,50(2):174~188.
[13] Kitagawa G.Monte Carlo filter and smoother for non-Gaussian nonlinear state space models[J].Journal of Computational and Graphical Statistics,1996,5(1):1~25.
[14] Bailey T,Nieto J,Nebot E.Consistency of the FastSLAM algorithm[A].Proceedings of the IEEE International Conference on Robotics and Automation[C].Piscataway,N J,USA:IEEE,2006.424~429.
[15] 陈得宅,赵春霞.一种改进遗传算法件能的方法研究[J].南开大学学报(自然科学版),2005,38(6):84~88.Chen De-bao,Zhao Chun-xia.A method to improve performance of genetic algorithm[J].Acta Scientiarnm Naturalium Universitatis Nankaiensis,2005,38(6):84~88.
[16] 厉茂海,洪炳熔,罗荣华.用改进的Rao-Blackweilized粒子滤波器实现移动机器人同时定位和地图创建[J].吉林大学学报(工学版),2007,37(2):401~406.Li Mao-hal,Hong Bing-rong,Luo Rong-hua.Improved RaoBlackwellized particle filters for mobile robot simultaneous localization and mapping[J].Journal of Jilin University (Engineering and Technology Edition),2007,37(2):401~406.