2019 International Conference on Power Electronics Applications and Technology in Present Energy Scenario (PETPES) | 2019
Vehicle Trajectory Prediction using Non-Linear Input-Output Time Series Neural Network
Abstract
Future location prediction of the target vehicle and the nearby object is one of the main challenges in automotive technologies for enhancing the road safety. Conventional approaches utilize kinematic motion models, for long duration prediction, the results obtained from these models are not reliable. In this study the trajectory prediction of a target vehicle is achieved using non-linear input-output time series neural network. The data set were collected from the CarMaker by creating various scenarios. The Artificial Neural Network (ANN) is trained by using neural network time-series tool box, and the developed model is evaluated to accomplish most precise and ideal model in prediction of longitudinal and lateral trajectory of the target vehicle.