No abstract

No abstract

No abstract

The development of autonomous vehicles is one of the most active research areas in the automotive industry. The objective of this study is to present a concept for analysing a vehicle’s current situation and a decision-making algorithm which determines an optimal and safe series of manoeuvres to be executed. Our work focuses on a machine learning-based approach by using neural networks for risk estimation, comparing different classification algorithms for traffic density estimation and using probabilistic and decision networks for behaviour planning. A situation analysis is carried out by a traffic density classifier module and a risk estimation algorithm, which predicts risks in a discrete manoeuvre space. For real-time operation, we applied a neural network approach, which approximates the results of the algorithm we used as a ground truth, and a labelling solution for the network’s training data. For the classification of the current traffic density, we used a support vector machine. The situation analysis provides input for the decision making. For this task, we applied probabilistic networks.

No abstract