Artificial Neural Network Modeling of an Autonomous Vehicle for Enhanced Lateral Position and Yaw Angle Prediction

Abstract

A key factor driving the development of autonomous vehicles (AVs) is safety. AVs should be able to navigate autonomously by generating appropriate waypoints and following them instantaneously. Thus, an effective path planning and trajectory tracking control system should be developed. Predicting the AV’s future path is an important issue due to its importance in trajectory tracking approaches. This paper proposes a novel neural network (NN) model for AV trajectory forecasting based on the applied steering angle. Predicting future lateral position and yaw angle before applying the steering angle guarantees that the optimal steering angle is used to minimize error. The NN model has been tested on different driving scenarios and simulation results validate the suggested NN model’s effectiveness.