Lane Change Assistance with AFS

Lane Change Assistance with AFS

For this research we considered a vehicle equipped with an Electric Power Assisted Steering (EPAS) and an Active Front Steering (AFS) unit. The steering assistance system uses the EPAS and the AFS to command a hand wheel torque feedback which can autonomously steer the vehicle or assist the driver. It also uses the AFS to superimpose an additional angle to the steering column. This allows control of the hand wheel rotation at different input feedback torque values.

An MPC is used to compute the optimal torque feedback input to the hand wheel along with the commanded AFS superposition angle in both autonomous and assisted driving scenarios. The MPC employs a model of the steering dynamics and a driver steering behavior in order to obtain lesser hand wheel rotation during a same lane change maneuver as compared to a conventional EPAS system without AFS.

We have validated the proposed approach by performing simulations for both autonomous and driver assisted scenarios. The results show significant decrease in hand wheel rotation due to the presence of AFS, while successfully performing the lane change. Further experiments are to be conducted on a test vehicle. 

Adaptive Learning Model Predictive Control for Autonomous Racing

Adaptive Learning Model Predictive Control for Autonomous Racing

Distributed Bayesian Filters for Multi-Vehicle Network

Distributed Bayesian Filters for Multi-Vehicle Network