2016 · 21 citations · 16 references
EngineeringComputer ScienceRobotics EngineeringMarkov Decision ProcessesPath PlanningRoboticsAutonomous NavigationMulti-robot LearningNavigation PlanningPolicy LearningReinforcement LearningRobot NavigationMobile Robot NavigationMulti-agent Path PlanningMulti-robot Path PlanningReinforcementReinforcement Learning For NavigationReinforcement Learning For RoboticsRobot Path PlanningRobotic Path Planning
This paper presents a direct heuristic dynamic programming (HDP) based on Dyna planning (Dyna_HDP) for online model learning in a Markov decision process. This novel technique is composed of HDP policy learning to construct the Dyna agent for speeding up the learning time. We evaluate Dyna_HDP on a differential-drive wheeled mobile robot navigation problem in a 2D maze. The simulation is introduced to compare Dyna_HDP with other traditional reinforcement learning algorithms, namely one step Q-learning, Sarsa (λ), and Dyna_Q, under the same benchmark conditions. We demonstrate that Dyna_HDP has a faster near-optimal path than other algorithms, with high stability. In addition, we also confirm that the Dyna_HDP method can be applied in a multi-robot path planning problem. The virtual common environment model is learned from sharing the robots' experiences which significantly reduces the learning time.
16
Reinforcement Learning: An Introduction
Richard S. Sutton, Andy Barto · IEEE Transactions on Neural Networks · 1998 · 27.1K citations
Learning to Predict by the Methods of Temporal Differences
Richard S. Sutton · Machine Learning · 1988 · 4K citations · Full text
Danil Prokhorov, Donald C. Wunsch · IEEE Transactions on Neural Networks · 1997 · 1.2K citations