Why RRT planner doesn't reach the goal state?
10 visualizaciones (últimos 30 días)
Mostrar comentarios más antiguos
Kseniya Kalita
el 6 de Jul. de 2020
Comentada: Kseniya Kalita
el 7 de Jul. de 2020
Hello everyone,
I am trying to understand the example https://de.mathworks.com/help/nav/ref/plannerrrtstar.html.
It is about motion planning using RRT algorithm.
There are start and goal positions
start = [0.5, 0.5 0];
goal = [2.5, 0.2, 0];
could you please explain me why the red line doesn't reach the goal state exactly?
If to run programm in Matlab in SoInInfo it is indicated that the plan is found.
0 comentarios
Respuesta aceptada
Cam Salzberger
el 6 de Jul. de 2020
Hello Kseniya,
The plannerRRTStar class has a GoalReachedFcn that checks if the state of one of the RRT nodes is sufficiently close to the goal to call it "complete". There needs to be some level of tolerance, since in the continuous state space, the robot will never exactly reach the goal. The default GoalReachedFcn is nav.algs.checkIfGoalIsReached, which uses a tolerance of 0.5 meters, and the chart shows the robot reaching within that tolerance.
If you want to use a more precise GoalReachedFcn, you can easily implement your own. The code in nav.algs.checkIfGoalIsReached is very simple, so it'd be easy to copy it and modify the tolerance distance to be smaller.
-Cam
Más respuestas (0)
Ver también
Categorías
Más información sobre Motion Planning en Help Center y File Exchange.
Productos
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!