Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
SOE: Fachverband Physik sozio-ökonomischer Systeme
SOE 6: Poster
SOE 6.8: Poster
Montag, 20. März 2017, 17:00–20:00, P2-OG4
Lane Change Prediction in an Urban Area — •Karoline Griesbach1 and Karl Heinz Hoffmann2 — 1Institute of Physics, Technische Universität Chemnitz, 09107 Chemnitz, Germany, Telephone: +49 371 531 35456 — 2Institute of Physics, Technische Universität Chemnitz, 09107 Chemnitz, Germany, Telephone: +49 371 531 35456
The prediction of the lane change and its integration in advanced driving assistance systems can reduce traffic accidents. In the article a neural network for lane change prediction will be discussed. The neural network was implemented with three learning rules: delta rule, backpropagation and backpropagation with momentum. The prediction of right and left lane changes were considered. The input data was provided by a Naturalistic Driving study and divided into a training set and a validation set. The best prediction was achieved for the left lane change with a neural network with backpropagation (tpr = 72.09%, fpr = 0.00%). The prediction of the right lane change was not successful.