Search

Path following and lateral stability are crucial issues for autonomous vehicles. Moreover, these problems increase in complexity when handling articulated heavy-duty vehicles due to their poor manoeuvrability, large sizes and mass variation. In addition, uncertainties on mass may have the potential to significantly decrease the performance of the system, even to the point of destabilising it. These parametric variations must be taken into account during the design of the controller. However, robust control techniques usually require offline adjustment of auxiliary tuning parameters, which is not practical, leading to sub-optimal operation. Hence, this paper presents an approach to path-following and lateral control for autonomous articulated heavy-duty vehicles subject to parametric uncertainties by using a robust recursive regulator. The main advantage of the proposed controller is that it does not depend on the offline adjustment of tuning parameters. Parametric uncertainties were assumed to be on the payload, and an $\mathcal{H}_{\infty}$ controller was used for performance comparison. The performance of both controllers is evaluated in a double lane-change manoeuvre. Simulation results showed that the proposed method had better performance in terms of robustness, lateral stability, driving smoothness and safety, which demonstrates that it is a very promising control technique for practical applications., 19 pages, 13 figures, preprint submitted to the Control Practice Engineering

Omnidirectional vision systems can provide images with a 360° of field of view. For this reason, they can be useful to robotic applications such as navigation, teleoperation and visual servoing. An effective way to construct this type of vision system is by combining lenses and mirrors together resulting on a system that does not require the movement of the camera to the direction of attention of the robot. A typical construction is by mounting a convex mirror in front of a camera aligning the center of the mirror with the optical axis of the camera. The most common convex mirror shapes used are conic, parabolic, hyperbolic and spherical. In this work we present two types of mirror that were constructed: a spherical mirror, used in the initial tests, and a hyperbolic, used for actual robot tasks. The hyperbolic mirror was manufactured using an ultra-precision CNC machine. Additionally, a software was developed to create panoramic and perspective images from the image acquired by the system. This work shows the development of an omnidirectional vision system, presenting the formulation used to determine a suitable mirror shape, the mechanical solutions used to build a fully operational system, and the results of the developed algorithm.