LaValle Planning Algorithms, chs 10, 11, 12

Ch 10

  • This basically all RL so little new here
  • (428) Nondeterministic Dijkstra

Ch 11

  • Covers POMDPs, information spaces/belief states, since its all impossible I went through this quickly as well
  • (477) More compact (lossy) belief states
  • (483) NFAs
  • (510) Sample-based approaches, particle filtering

Ch 12

  • Prev ch. covered information spaces, this covers planning in them
  • (524) Constraints on the type of partial observability can lead to more tractable planning
  • (528) Localization
  • (545) Mazes can be searched in space log in # tiles, need to know how far from top agent is and its orientation, does wall following with small transforms done to the map
  • (549) D* is a backwards variant of Dijkstra’s, calculates cost-to-go starting from goal
  • (550) Bug algos for planning in unknown continuous environments
    • Traces around walls, assumes only very local visibility
    • Boundary of each obstacle is followed no more than 3/2 times
    • May only be 2D again
  • (554) Competitive rations can be used to determine efficiency of an online algo, basically comparison is between the online algo and one that already has all the information
  • (555) Optimal navigation w/ver  limited sensing (may have dim restriction)
  • (561) SLAM/EM
  • (571) Cases where having less information can make planning simpler
Tagged ,

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: