Multi-sensor tracking of a vehicle on a grid

Considerable work has been done on model-based, multi-sensor signal processing algorithms for estimating the location and the motion mode of a mobile vehicle. Hybrid models provide the best framework for fusion of disparate sensor measurements. But abrupt sensor nonlinearities and rigid motion constraints require more detailed analysis. This paper presents a multiple-model estimator that provides high quality location and uncertainty estimates for a vehicle following a mapped road grid. An example illustrates the power of the estimator.