METHOD FOR PREDICTING THE INTENTIONAL CUT-IN OF NEIGHBORING VEHICLES USING ARTIFICIAL INTELLIGENCE

摘要

The present invention relates to a method for predicting cut-in intention of surrounding vehicles using artificial intelligence, and more particularly, to a prediction method for determining the cut-in intention of surrounding vehicles by using PreScan and Vissim to acquire data and applying 3D CNN and LSTM. The present invention includes the steps of generating data to which a driving model is applied by applying a vehicle simulator, PreScan, and a traffic simulator, Vissim, to surrounding vehicles of an autonomous vehicle; Pre-processing the data by arranging the data to which the driving model for the autonomous vehicle and surrounding vehicles is applied, including the autonomous vehicle as central data, as information on 9 vehicles, and forming a data states buffer ; And determining a cut-in intention by applying a long short term memory (LSTM) by applying a 3D CNN to the data state buffer to calculate a correlation based on driving information between the autonomous vehicle and the surrounding vehicle. It provides a method of predicting the intention to cut-in surrounding vehicles using artificial intelligence. According to the present invention, there is an advantage of being able to quickly and quickly reduce the possibility of an accident that may occur in a situation in which an autonomous vehicle changes lanes by determining the intention of a nearby vehicle.