Maschinelles Lernen zur Optimierung einer autonomen Fahrspurführung


In this work a model-free path tracking method was developed. The method considers the vehicle kinematics, the path geometry and the delayed control actions in order to determine the optimal control action. The predictive path tracking method applies the Neural Fitted Q Iteration algorithm for training a neural network. The trained neural network provides an approximation of the Q-function so that the best control action for the current vehicle state can be determined.