A novel approach is used for trajectory planning and model predictive control design for automated driving in certain critical environments. An active safety system is designed to avoid the potential collisions and generate a reference trajectory by using polynomial parameterization. Non-linear optimal control technique is used to calculate the coefficients of the polynomial within the boundary conditions specified. The path planning algorithm is designed in a way to generate a trajectory maintaining a safe distance from the obstacle when the obstacle moves from one lane to other or is in the same lane. Model predictive controller (MPC) is designed to stabilize the dynamic bicycle model along the desired path by satisfying the constraints, weighting functions and the reference trajectory given. The MPC controller generates a steering angle input to the system to control the lateral position of the vehicle. The model is tested for various conditions of the obstacle and the vehicle model, such as when the obstacle is moving with constant velocity, variable velocity, is in static position and moving at any random positions between the two lanes. The simulation results demonstrate the performance of the controller and the path planning algorithm with obstacle avoidance.
A model predictive control approach for autonomous active safety systems based on trajectory planning and collision avoidance
2018-05-01
Sonstige
Elektronische Ressource
Englisch
DDC: | 629 |
Predictive Safety Systems - from Convenience to Active Collision Avoidance
British Library Conference Proceedings | 2004
|Bézier curve‐based trajectory planning for autonomous vehicles with collision avoidance
Wiley | 2020
|Bézier curve-based trajectory planning for autonomous vehicles with collision avoidance
IET | 2021
|