Nowadays, in the car industry, the driving assistance technologies have more role, than before. These are include lots of highly importance conformity and safety functions, with the traffic situation recognition and evaluating. Among other things, witch traffic maneuvers are under performed, and what will they intend to do?
The situation analysis is based on uncertain data, and the most of the cases it requires the uncertain decision systems. It seems like the best choice is using (graphical) probabilistic models for the solution.
At the beginning of my thesis, I become familiar with the basics of the probabilistic models, and then I describe my development. The lane changing recognition, and its forecasting method is written in MATLAB environment. After the signals of the radar and camera systems is preprocessed, and uploaded to the model, I evaluate tests and results.
That function can supply a basis for the software of quantity production cars, and provide information for the higher software layers.