Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.

  

Model Predictive Control for the Maneuvering of Cars

Author(s):

C. Ruch
Conference/Journal:

Master Thesis, HS13 (10246)
Abstract:

In this master project a model predictive control scheme for autonomous vehicles was developed. The scheme is designed for maneuvering a vehicle in low speed situations in the presence of input and state constraints. For those situations the correct modeling of the vehicle and the tyre - street interaction is equally important as the avoidance of obstacles. The vehicle is modeled with a dynamic bicycle model in a planar setting with no-slip conditions at the wheels imposing two nonholonomic constraints on the system. The static obstacles present in most low - speed maneuvers are modeled as an union of convex sets. This union restricts the set of admissible states for the vehicle in the optimization problem of the model predictive control scheme to a non-convex set. Those constraints are considered in the optimization problem by the computation of the signed distance between the vehicle and the individual obstacles. The signed distance function is included as a linear approximation in the optimization problem. An implementation of the control schme on the Berkeley Library for Optimization Modelling (BLOM) using an IPOPT nonlinear solver is presented. Simulation results for the exemplary situations parallel parking and lot parking are shown. Experimental results for a simple low speed maneuver on a Hyundai Grandeur test vehicle using a dSpace implementation and an NPSOL nonlinear solver are presented as well. While general performance including passenger comfort of the control scheme is satisfying, the underlying low level control loops for steering and acceleration have to be reevaluated.

Year:

2014
Type of Publication:

(12)Diploma/Master Thesis
Supervisor:

F. Borrelli

File Download:

Request a copy of this publication.
(Uses JavaScript)
% Autogenerated BibTeX entry
@PhdThesis { Xxx:2014:IFA_4785
}
Permanent link