On Warning For Dangerours Vehicle Lane Changes Considering Driving Intentions

Vehicle lane change happens frequently on highway, which involves both the longitudinal and lateral control of the vehicle, both acceleration and deceleration processes, and surveillances of both the original and targeted lanes. Therefore, it poses relatively high requirement on drivers’attention as well as driving skills. Vehicle lane change pre-warning systems, which play an important role in advanced driving assistance systems, can assess lane change situations automatically and give warnings when potential dangers appear. As is known, different driving intentions will lead to different driving behaviors, and further result in different lane change situations. So a reliable pre-warning system, similar to the situation evaluation processes of a human driver, should consider not only the vehicle states, but also the intentions of the vehicles around.Driving intentions of the four involved vehicles in the lane change scenario, i.e., the leading and following vehicles in both the original and targeted lanes can be interesting. Intentions of each of these four vehicles are analyzed and corresponding quantitative driving intention values are obtained taking advantage of fuzzy logic which mimics human intuitive inference system. The concept of time headway is adopted to decide the safety lane change conditions based on an analysis of the kinematics during lane change; furthermore, different critical levels are rated. Time headway is expressed as the function of the initial relative spaces and velocities between the subject vehicle and the four involved vehicles. Based on the safety lane change model and the induced driving intentions, an improved safety lane change model with consideration of driving intentions is established. Both the original and improved models are simulated in Matlab. Different lane change scenarios under different initial relative spaces and velocities are set up in CarSim and the lane change data is then imported into a Matlab/GUI which is designed based on the proposed lane change model. Effects of driving intentions on the lane change pre-warning system are analyzed in detail. Simulation results show that the utilization of time headway instead of Minimum Safety Spacing in the lane change model further enhances safety. Driving intentions deduced by fuzzy logic system have a significant influence on the lane change situation and thus the pre-warning results. Especially, the deceleration intention of the following vehicle in the desired lane could cooperate with the subject vehicle in the process, which will in turn expand the safety area and reduce the frequency of unnecessary warnings.In summary, the safety lane change model considering driving intention proposed in this thesis can improve the accuracy of lane change situation assessment and further enhance the pre-warning performance. It can therefore be of significance both in theory and practice. This research provides a foundation for the potentially wide use of lane change assistance systems.