6.4210 Fall 2023 Lecture 12: Motion Planning- Sampling Based and Global Optimization

แชร์
ฝัง
  • เผยแพร่เมื่อ 17 พ.ย. 2024

ความคิดเห็น • 2

  • @xiaoyuwang6841
    @xiaoyuwang6841 11 หลายเดือนก่อน

    Hi Professor Tedrake, I have a question regarding RRT algorithm. If there is not a biased small probability towards the goal when generating the random samples, is the algorithm guarantee the goal will be reached? If the goal is on an interpolated line from the tree to a sample, is it considered as the goal has been reached? Thank you very much!

    • @rtedrake
      @rtedrake 11 หลายเดือนก่อน

      I suspect that one could claim rigorously that some finite volume region around the goal would be reached with probability one as the number of sample goes to infinity. But the goal selection bias would be required to make convergence to a point possible.