This work evaluates a real-time algorithm to localize a vehicle on a highway in the direction of travel without the use of GPS. The algorithm uses a particle filter to estimate vehicle position along a map of road grade using real-time pitch measurements from an in-vehicle pitch sensor as the input. Experiments over 60 kilometers along Interstate I-80 and US Route 220 in Pennsylvania are used to demonstrate the algorithm, observe the speed of convergence, and evaluate several methods of implementation. The results indicate that the method can localize a vehicle with a position accuracy of 5 meters after traveling about 1 kilometer within the 60 kilometer map.
This work develops a particle filter algorithm to localise a vehicle in the direction of travel without the use of GPS. The inputs to the algorithm include a terrain map of road grade, pitch measurements from an in-vehicle pitch sensor, and wheel odometry. Simulations and experiments at The Thomas D. Larson Transportation Institute test track are used to demonstrate the algorithm, observe the speed of convergence, and to determine key parameters for practical implementation. The results indicate that the method can quickly localise a vehicle with 1 m accuracy or better. Experiments over 5 km along Highway 322 in State College, Pennsylvania, were also used to demonstrate the algorithm.
This work develops an algorithm for estimating the lateral lane index of road vehicles on multi-lane roadways by correlating vehicle attitude measurements to terrain maps of the individual lanes of travel. To localize a vehicle, a Bayesian belief algorithm and a particle filter algorithm are described and applied off-line using data collected from two lanes along a local highway. Results demonstrate that terrain-based algorithms are capable of measuring lane index. Because these measurements are immune to lighting conditions, this solution is a good complement to existing lane-detection camera systems.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.