Evolving the Controller of Automated Steering of a Car in Slippery Road Conditions

Abstract

Abstract:The most important characteristics of autonomous vehicles are their safety and their ability to adapt to various traffic situations and road conditions. In our research, we focused on the development of controllers for automated steering of a realistically simulated car in slippery road conditions. We comparatively investigated three implementations of such controllers: a proportional-derivative (PD) controller built in accordance with the canonical servo-control model of steering, a PID controller as an extension of the servo-control, and a controller designed heuristically via the most versatile evolutionary computing paradigm: genetic programming (GP). The experimental results suggest that the controller evolved via GP offers the best quality of control of the car in all of the tested slippery (rainy, snowy, and icy) road conditions.

PD Steering Controller Utilizing the Predicted Position on Track for Autonomous Vehicles Driven on Slippery Roads

“…For the experiments, a racing model of a rear-wheel-drive car of the Mercedes brand was used, and a list of its parameters is shown in Table 1 below, and Figure 1 shows its view during the races. As we mentioned earlier [6], the choice of TORCS over different alternatives as a simulator in our experiments was also determined by its computational efficiency, safety, and the availability of its source code.…”

“…The peculiarity of this rotation is that it does not contain a transition spiral curve part between the straight and the rotation sections. A spiral curve is As we mentioned earlier [6], the choice of TORCS over different alternatives as a simulator in our experiments was also determined by its computational efficiency, safety, and the availability of its source code.…”

“…In our earlier studies, we showed that the canonical well-known servo control ( Figure 4) is a variation of the PD controller. To avoid repetition, we give here only the main statements; the details can be found in our paper [6].…”

“…The optimization of these parameters by human experts often requires extensive knowledge in both the control theory and vehicle dynamics. Automated tuning of the parameters, on the other hand, might require applying heuristic approaches that are notorious with their long runtime even when using a significant computational power [4,5].…”

“…In order to address the above-mentioned challenges of the canonical PD controllers, in which the control output is calculated as a weighted sum of the control errors and their derivatives, in our previous research we proposed a PID steering controller featuring an arbitrary (rather than an additive) internal structure, developed heuristically via genetic programming [6].…”