TY - CONF T1 - Optimal estimation techniques to reduce false alarms in railway obstacle detection T2 - IEEE International Conference on Industrial Technology, 2005. ICIT 2005. Y1 - 2005 A1 - J.J. Garcia A1 - Cristina Losada-Gutiérrez A1 - Felipe Espinosa A1 - Ureña, J. A1 - Álvaro Hernández A1 - Manuel Mazo A1 - Carlos De Marzziani A1 - Ana Jimenez A1 - Fernando J. Álvarez KW - collision avoidance KW - filtering theory KW - Hinfin control KW - Kalman filters KW - principal component analysis KW - railway accidents AB - This work presents some methods to reduce false alarms in railway obstacle detection. The sensorial system is based on one barrier of infrared emitters and another of receivers, placed on opposing sides of the railway. Obstacle detection is achieved by the lack of reception in the detectors. On the one hand, the efficiency of the system is achieved with the geometrical distribution of the sensorial system and the codification used in the emitting and receiving stages. On the other hand, optimal estimation techniques have been proposed to avoid false alarms, based on Kalman and Hinfin filtering. Principal component analysis is developed to validate the obstacle detection, and to improve the accuracy of the system. A high reliability under adverse conditions is obtained with the barrier, it being possible to detect the presence of obstacles, and to report on their position. JF - IEEE International Conference on Industrial Technology, 2005. ICIT 2005. CY - Hong Kong SN - 0-7803-9484-4 UR - http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=1600682&isnumber=33650 ER -