Combining Evolution and Training in a Robotic Controller for Autonomous Vehicle Navigation
This work investigates a robotic control system designed to autono-mously navigate a vehicle in urban environments. Our approach is based on the use of two Artificial Neural Networks (ANNs), one is trained for image processing with road recognition and template matching and the second is evolved for navigation control. This paper focuses on experiments and evaluations using a Genetic Algorithm (GA) to evolve the ANN responsible to provide steering and velocity control to the autonomous vehicle.
- 2.Souza, J.R., Pessin, G., Eboli, G.B., Mendes, C.C.T., Osório, F.S., Wolf, D.F.: Vision and gps-based autonomous vehicle navigation using templates and artificial neural networks. In: The 27th ACM Symposium on Applied Computing, pp. 1008–1013 (2012)Google Scholar