Learning Coordinated Behaviors for Control of a Simulated Race Car

category: Control, navigation and planning This work is not in submission or press to any other conference. Abstract The demands of rapid response and the complexity of many environments make it diicult to decompose, tune and coordinate re-active behaviors while ensuring consistency. We hypothesize that complex behaviors should be decomposed into separate behaviors resident in separate networks, coordinated through a higher level controller. To explore these issues, we have implemented a neu-ral network architecture as the reactive component of a two layer control system for a simulated race car. By varying the architecture , we tested whether decomposing reactivity into separate behaviors leads to superior overall performance and learning convergence. Based on these results, we further modiied the architecture to produce a race car that is competitive with publicly available solutions.