Learning Autonomous Helicopter Flight with Evolutionary Reinforcement Learning
In this paper we present a method to obtain a near optimal neuro-controller for the autonomous helicopter flight by means of an ad hoc evolutionary reinforcement learning method. The method presented here was developed for the Second Annual Reinforcement Learning Competition (RL2008) held in Helsinki-Finland. The present work uses a Helicopter Hovering simulator created in the Stanford University that simulates a Radio Control XCell Tempest helicopter in the flight regime close to hover. The objective of the controller is to hover the helicopter by manipulating four continuous control actions based on a 12-dimensional state space.
KeywordsReinforcement Learning Evolutionary Computation Autonomous Helicopter
Unable to display preview. Download preview PDF.
- 2.Moriarty, D.E., Schultz, A.C., Grefenstette, J.J.: Reinforcement learning through evolutionary computation (1999)Google Scholar
- 3.Koppejan, R., Whiteson, S.: Neuroevolutionary reinforcement learning for generalized helicopter control. In: GECCO 2009: Proceedings of the Genetic and Evolutionary Computation Conference (to appear, July 2009)Google Scholar
- 4.Ng, A.Y., Kim, H.J., Jordan, M.I., Sastry, S.: Autonomous helicopter flight via reinforcement learning. In: Thrun, S., Saul, L.K., Schölkopf, B. (eds.) NIPS. MIT Press, Cambridge (2003)Google Scholar
- 5.Jose Antonio Martin, H., de Lope, J., Santos, M.: Evolution of neuro-controllers for multi-link robots. In: Innovations in Hybrid Intelligent Systems. Advances in Soft Computing, vol. 44, pp. 175–182 (2008)Google Scholar