Evaluation of Non-GPS Train Localization Schemes Using a Commodity Smartphone with Built-In Sensors

摘要

In this paper, we compare performances of train localizationschemes by the dynamic programming of various sensor informationobtained from a smartphone attached to a train, and further discuss themost superior sensor information and scheme in this localization system.First, we compare the localization performances of single sensor informationschemes, such as 3-axis acceleration information, acoustic information,3-axis magnetic information, and barometric pressure information. Thesecomparisons reveal that the lateral acceleration information input schemehas the best localization performance. Furthermore, we optimize each datafusion scheme and compare the localization performances of the data-fusionschemes using the optimal ratio of coefficients. The results show that thehybrid scheme has the best localization performance, with a root meansquared error (RMSE) of 12.2 m. However, there are no differences betweenthe RMSEs of the input fusion scheme and 3-axis acceleration inputscheme in the most significant three digits. Consequently, we conclude thatthe 3-axis acceleration input fusion scheme is the most reasonable in termsof simplicity.