Name
Abstract | ||
|---|---|---|
Motion-prediction algorithms for vehicles often employ historical behavior of a vehicle, rely on the Markov property of the underlying system, and predict the future behavior of the vehicle. However, the Markov property alone may lead to conservative predictions and heavy computational burden. To overcome these drawbacks, this paper develops a method that uses the notion of similarity among vehicle trajectories. As traffic rules and driver intentions restrict the motions of a vehicle, the behavior of a road vehicle is typically similar to that of other vehicles. We hypothesize that if the motion of any two vehicles was similar in the past for a sufficiently long time span, then it is likely that it will be similar in the future. This paper introduces an algorithm that exploits this hypothesis to develop prediction methods, and from the results of numerical simulations, it verifies the effectiveness of the algorithm. |
Year | DOI | Venue |
|---|---|---|
2017 | 10.23919/ACC.2017.7962970 | 2017 American Control Conference (ACC) |
Keywords | DocType | ISSN |
similarity-based vehicle-motion prediction,Markov property,conservative predictions,traffic rules,vehicle trajectories,road vehicle behavior,numerical simulations | Conference | 0743-1619 |
ISBN | Citations | PageRank |
978-1-5090-4583-9 | 0 | 0.34 |
References | Authors | |
5 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Kazuhide Okamoto | 1 | 3 | 2.48 |
| Karl Berntorp | 2 | 26 | 16.30 |
| Stefano Di Cairano | 3 | 309 | 44.69 |