Strategy Synthesis for POMDPs in Robot Planning via Game-Based Abstractions.
Leonore WintererSebastian JungesRalf WimmerNils JansenUfuk TopcuJoost-Pieter KatoenBernd BeckerPublished in: IEEE Trans. Autom. Control. (2021)
Keyphrases
- motion planning
- mobile robot
- belief space
- partially observable markov decision processes
- belief state
- goal directed
- world model
- partially observable
- learning process
- vision system
- robot arm
- human robot interaction
- reinforcement learning
- continuous state
- planning under uncertainty
- robotic systems
- learning tools
- path planning
- finite state
- collision free
- mechanical devices
- point based value iteration
- stochastic domains
- navigation tasks
- high level
- markov decision problems
- multiple robots
- dynamic programming
- optimal policy
- humanoid robot
- robot manipulators
- admissible heuristics
- degrees of freedom
- ai planning
- multi robot
- predictive state representations
- autonomous robots